JP3947421B2 - Conversation control system, conversation control method, program, and recording medium recording program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a conversation control system, a conversation control method, a program, and a recording medium on which a program is recorded, which obtains optimum answer contents for the input information based on input information input from a user.
[0002]
[Prior art]
In the conventional conversation control system, since the answer contents stored in advance corresponding to the utterance contents are output based on the utterance contents from the user, the user can use the answer contents answered from the conversation control system. I was able to enjoy the feeling of talking with other users in a pseudo manner.
[0003]
[Problems to be solved by the invention]
However, in the above conversation control system, the utterance content (input information) from the user is collated with the utterance content (phrase) group of the user stored in advance, and matches the input information from the phrase group. Since it searches for a phrase and outputs the response content associated with the searched phrase, if it does not store a large number of phrases that are as close as possible to the above input information, it will output the response content that is more appropriate for the input information. I couldn't.
[0004]
On the other hand, if the developer who constructs the conversation control system stores in the conversation control system in advance a huge number of phrases that are as close as possible to the input information and the answer contents associated with the large number of phrases, the conversation control system Can search for a phrase that matches the input information from a huge group of phrases stored in advance, and output the answer contents associated with the searched phrase, so that the answer is more suitable for the input information. The contents can be output.
[0005]
However, a developer who builds a conversation control system can search for a phrase that matches the input information using the conversation control system, and the user will have a huge phrase group (a phrase group near infinity). Must be stored in the conversation control system in advance, and it is difficult to construct a conversation control system that outputs answer contents more suitable for input information.
[0006]
Therefore, the present application has been made in view of the above points, and each morpheme constituting input information input from the user is specified, and the semantic content grasped from each specified morpheme is analyzed and analyzed. By obtaining pre-prepared answer contents associated with semantic contents, the optimum answer contents corresponding to the input information from the user are output, and it is appropriate even if each answer content stored in advance is reduced. It is an object of the present invention to provide a conversation control system, a conversation control method, a program, and a recording medium on which the program is recorded, which can output answer contents.
[0007]
[Means for Solving the Problems]
The invention according to the present application specifies a character string indicating input information based on input information input from a user, and based on the specified character string, at least one morpheme consisting of a minimum unit constituting the character string. As the first morpheme information, and based on the specified character string, specify the type information indicating the tense constituting the character string or the type of request for a certain matter, one character, a plurality of character strings or these The second morpheme information indicating a combination of morphemes is associated with a plurality of response contents to the user, and each answer content is associated with an answer type indicating the type of the answer contents. A plurality of information is stored in advance, the extracted first morpheme information and each second morpheme information are collated, and second morpheme information including the first morpheme information is searched from each second morpheme information. Based on the second morpheme information, each answer type associated with the second morpheme information is compared with the specified type information, and the answer type matching the type information is searched from each answer type. The answer content associated with the answer type is acquired based on the answer type.
[0008]
According to the invention according to the present application, each conversation control device is associated with the second morpheme information based on the second morpheme information including the first morpheme information constituting the input information input from the user. The response type and the specified type information (the tense that constitutes the character string or the type of input information consisting of requests for a certain matter) are collated, and the response type that matches the type information is searched from each response type. Since the response content associated with the retrieved response type can be acquired, the conversation control device can determine the type of input information input by the user, particularly a tense that constitutes a character string or a request for a certain matter. Based on the type, the corresponding answer content can be acquired, and as a result, the answer content suitable for the input information from the user can be acquired.
[0009]
In the invention according to the present application, the second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof is associated with a plurality of response contents to the user, and each answer content corresponds to the input information. Priorities for selection as response contents are associated with each other, a plurality of second morpheme information is stored in advance, and a character string indicating input information is specified based on input information input from a user. , Based on the identified character string, extract at least one morpheme consisting of the minimum unit constituting the character string as first morpheme information, collate the extracted first morpheme information and each second morpheme information, Search for second morpheme information including first morpheme information from each second morpheme information, and second morpheme including first morpheme information from each second morpheme information based on the searched search results. Ranking is performed according to the frequency of search information, and based on the searched second morpheme information, each priority order associated with the second morpheme information is compared with the ranked rank. A priority order that matches the rank is acquired from each priority order, and answer contents associated with the priority order are acquired based on the acquired priority order.
[0010]
In addition, when the said frequency searched is low, it is desirable to rank to a low rank. Moreover, when the said frequency searched is high, it is desirable to rank to a high rank.
[0011]
According to the invention according to the present application, the conversation control device ranks according to the frequency of searching for the second morpheme information including the first morpheme information from each second morpheme information, Based on the searched second morpheme information, each priority order associated with the second morpheme information is compared with the ranked rank, and the priority order that matches the rank is obtained from each priority order. Since the answer contents associated with the priority order can be acquired based on the acquired priority order, the conversation control device refers to the rank ranked with reference to the search result (search process). If it is an emotion level held for a user, it is possible to obtain the response content associated according to the magnitude of this emotion level, and in this case, the corresponding response content according to the emotion level held for the user To get It can be.
[0012]
In addition, if the answer contents that are of interest to the user are provided according to the priority, the conversation control device is particularly useful for the user when the rank ranked by referring to the search result is large. You can get the answers you are interested in.
[0013]
As a result, for example, if the user does not input input information including first morpheme information that can be searched by the conversation control device, the conversation control device searches for second morpheme information that matches the first morpheme information. In this case, the rank is set to a low rank, so that it is impossible to acquire information that is interesting to the user as described above.
[0014]
For this reason, in order to acquire information that is of interest to the user, the user must input input information that is not set low in the rank ranked by the conversation control device. It is possible to acquire specific information while tasting the feeling of having a conversation with care.
[0015]
Moreover, since the developer who develops the conversation control device can specify, for example, ranking as an emotion level based on the search history performed between the first morpheme information and each second morpheme information, It is no longer necessary to develop a complicated algorithm or neural network for causing the conversation control device to calculate the emotion level held for the user, and consequently, the development time for developing the device for achieving the present invention is reduced. It can be shortened.
[0016]
In the invention according to the present application, a plurality of second morpheme information is associated with a discourse range indicating morphemes that are related to input information that will be input from the user or to the content of an answer to the user, and each second morpheme A plurality of response contents to the user are associated with the information, and each answer content is associated with a priority order for being selected as a response content for the input information, and a plurality of the conversation ranges are stored in advance. The first morpheme information is compared with each discourse range, the discourse range including the first morpheme information is searched from each discourse range, and the first discourse range indicating the previous discourse range searched And the second discourse range indicating the later discourse range, and determines whether the second discourse range is related to the first discourse range, and ranks based on the determination result is there. In addition, when it is determined that the second discourse range is not related to the first discourse range, it is desirable to rank the lower rank.
[0017]
According to such an invention according to the present application, the conversation control device includes a first discourse range related to the first morpheme information searched first, a second discourse range related to the first morpheme information searched later, and Therefore, the conversation control device ranks the first morpheme information and each second morpheme information based on the search result, and ranks the ranked size. The answer contents associated with the matching priority order can be acquired, and as a result, the optimum answer contents for the input information from the user can be acquired.
[0018]
For example, if the second discourse range has no relationship with the first discourse range, the conversation control device is configured to input the current input information (input from the user) and previously input information. It is possible to determine that the content is completely different from that of the user, so the user can conclude that the input is inconsistent with respect to a certain topic, for example, set the rank to a lower rank. Can do.
[0019]
In this case, since the conversation control device has set the rank low, it can output the answer contents associated with the priority (low priority) matching the low rank, for example, the user is not particularly interested. Information (information with low priority) can be output as the response content.
[0020]
On the other hand, if the second conversation area has a relationship with the first conversation area, the conversation control device can input the current input information (input from the user) and the previously input information. Can be determined to have relevance, so the user can conclude that he / she has entered consistency in a topic, for example, the rank can be set to a higher rank. it can.
[0021]
In this case, since the conversation control device has set the rank higher, it can output the answer contents associated with the priority (high priority) matching the higher rank, for example, information that the user is particularly interested in (Information with high priority) can be output as the response content.
[0022]
In the invention according to the present application, the first morpheme information is composed of each morpheme associated with each attribute such as a subject case consisting of a main case and a target case consisting of a target case, and based on the extracted first morpheme information, Search for an attribute that does not contain a morpheme from the attributes that make up one morpheme information, add the morpheme that makes up the searched discourse range to the attribute based on the searched attribute, and configure the discourse range The morpheme is ranked based on the result of adding the morpheme.
[0023]
According to the invention according to the present application, when the morpheme of the discourse range is added to each attribute constituting the first morpheme information, the conversation control device is abbreviated as input information including the first morpheme information. Therefore, it can be determined that the user is inputting input information on the assumption of the discourse range to which the first morpheme information previously searched belongs.
[0024]
In this case, since the conversation control device can determine that the user is inputting input information on the assumption that the user has previously searched for the conversation range to which the first morpheme information belongs, the user can It can be determined that the input is consistent, and for example, the rank can be set to a higher rank.
[0025]
For this reason, the conversation control device can rank more appropriately based on the first discourse range and the second discourse range, and outputs the optimum answer content based on the appropriately ranked size. can do.
[0026]
The invention according to the present application stores a plurality of fixed contents for answering fixed contents in response to input information from a user, collates the extracted first morpheme information with each fixed contents, The fixed content including the first morpheme information is searched from the content, and the fixed content including the first morpheme information is searched from each fixed content based on the searched search result. It ranks according to the feature.
[0027]
According to the invention according to the present application, the conversation control device collates the first morpheme information and each fixed content, searches the fixed content including the first morpheme information from each fixed content, and performs this search. Since the ranking can be performed based on the result, the conversation control device searches, for example, the fixed content including the first morpheme information, and this fixed content is the content (return) to hear back the answer sentence from the conversation control device. In this case, since the user can determine that the input information for interrupting a certain topic is input, for example, the rank can be set to a lower rank.
[0028]
For this reason, the conversation control device can rank more appropriately based on the first morpheme information and the fixed content, and output the optimum answer content based on the appropriately ranked size. Can do.
[0029]
On the other hand, when the conversation control device cannot search the fixed content including the first morpheme information, the user does not input the input information for interrupting a certain topic, and the current topic is not input. On the other hand, since it can be determined that the answer is serious, the rank can be set to a higher rank, for example.
[0030]
For this reason, the conversation control device can rank more appropriately based on the first morpheme information and the fixed content, and outputs more optimal answer content based on the appropriately ranked size. can do.
[0031]
The invention according to the present application specifies type information indicating types such as affirmation, question, and repulsion constituting input information based on the specified character string, and when the specified type information includes repulsion, It is characterized by being ranked in a lower rank.
[0032]
According to the invention according to the present application, the conversation control device specifies type information indicating the type of input information, and when the specified type information is repelled, the conversation control device sets the rank to a lower rank. For example, when the user is repelling the response content output from the conversation control device, the device can set the rank low.
[0033]
For this reason, the conversation control device can rank more appropriately based on the type information. For example, if the rank is low, information that is not of interest to the user can be output as the response content.
[0034]
The invention according to the present application stores the acquired past response content, collates the acquired current first morpheme information with the past response content, and the current first morpheme information is included in the past response content. It is characterized by determining and ranking based on the determination result.
[0035]
According to the invention according to the present application, the conversation control device collates the current first morpheme information with the past answer content, and when the current first morpheme information is included in the past answer content, The conversation control device can conclude that the user is turning back with respect to the past response content (that the user is listening to the response content again).
[0036]
In this case, the conversation control device can conclude that the user has input the input information for interrupting a certain topic because the user is reversing the past answer contents, For example, the rank can be set to a lower rank.
[0037]
For this reason, the conversation control device can rank more appropriately based on the current first morpheme information and the past answer contents, and the optimum answer contents based on the appropriately ranked size Can be output.
[0038]
The invention according to the present application stores the acquired past first morpheme information, collates the acquired first first morpheme information with the past first morpheme information, It is characterized by determining whether it is contained in one morpheme information, and ranking based on a determination result.
[0039]
According to the invention according to the present application, the conversation control device collates the current first morpheme information with the past first morpheme information, and the current first morpheme information is included in the past first morpheme information. In this case, the user can simply determine that the input information input in the past is re-input regardless of the response content from the conversation control device.
[0040]
In this case, the conversation control device has nothing to do with the response content from the conversation control device, and the user simply inputs the same input information as the previous input information repeatedly. Ranks can be ranked, and information that is not of interest to the user can be output as answer contents.
[0041]
The invention according to the present application is characterized in that, when the ranked rank is the lowest, the response content is not acquired. In addition, a plurality of repulsion contents for repelling a certain matter are stored in advance, the extracted first morpheme information and each repulsion content are collated, and the repulsion contents including the first morpheme information from each repulsion content When the repelled content including the first morpheme information is searched when the ranked rank is the lowest, it is desirable not to acquire the response content.
[0042]
According to such an invention according to the present application, the conversation control device can prevent the response content from being acquired when the rank is set to the lowest rank. It can be used as a device that lets the user plant a sense of being.
[0043]
The invention according to the present application stores a plurality of apology details for apologizing for a certain matter in advance, collates the extracted first morpheme information with each apology content, and from each apology content, the first morpheme information If the apology content including the first morpheme information is searched when searching for the apology content including, and the response content is not acquired, the priority order is later determined based on the acquired priority order. It is characterized by acquiring the associated answer contents.
[0044]
According to the invention according to the present application, when the conversation control device performs a process of not acquiring the response content, when the apology content including the first morpheme information is searched, the acquired priority is later. Since it is possible to obtain the response content associated with the priority order based on the ranking, the conversation control device, when performing the process of not obtaining the response content, if the apology content is input from the user, It is possible to cancel the process of not obtaining the response content.
[0045]
As a result, the conversation control device can be used not only as a device that makes the user plant an angry sensation, but also as a device that allows the user to plant a sensation that makes the user feel angry. Can be used.
[0046]
The invention according to the present application relates to the morpheme according to the semantic content of the topic requested by the user, the emotion level held for the user with respect to a certain matter, or the type of statement, affirmative sentence, question sentence, repulsive sentence, etc. It is described in a language for hierarchically storing in a database. The language is preferably made up of DKML.
[0047]
As a result, the developer who develops the conversation control device can hierarchically construct the second morpheme information and the like for searching the answer content for the input information of the user using the language in the database. The conversation control device can search the database for second morpheme information related to the first morpheme information based on the first morpheme information corresponding to the user input information through a hierarchical procedure.
[0048]
That is, the conversation control device determines the hierarchy of the first morpheme information corresponding to the input information (for example, whether it is in a higher concept or lower concept with respect to the second morpheme information stored in the database). Appropriate second morpheme information can be searched from each second morpheme information stored in advance based on the determined hierarchy.
[0049]
For this reason, the conversation control device does not collate the first morpheme information consisting of user input information with the previously stored “all” second morpheme information step by step, The second morpheme information and the first morpheme information need only be collated, so that the second morpheme information that approximates the first morpheme information can be searched in a short time. As a result, the second morpheme information is set based on the search result. It is possible to quickly output the answer contents associated with the ranked rank.
[0050]
The second morpheme information including the first morpheme information (including the discourse range, fixed content, answer content, etc., hereinafter abbreviated as “second morpheme information etc.”) It means not only the case where the information is included as it is, but also the case where the second morpheme information or the like includes at least one morpheme constituting the first morpheme information.
[0051]
DETAILED DESCRIPTION OF THE INVENTION
[First embodiment]
(Basic configuration of conversation control system)
A conversation control system according to the present invention will be described with reference to the drawings. FIG. 1 is a schematic configuration diagram of a conversation control system having a conversation control apparatus 1 (ranked apparatus) according to the present embodiment.
[0052]
As shown in the figure, the conversation control device 1 includes an input unit 100, a speech recognition unit 200, a conversation control unit 300, a sentence analysis unit 400, a conversation database 500, an output unit 600, and a speech recognition dictionary storage. Part 700.
[0053]
In the present embodiment, for convenience of explanation, the description is limited to the user's utterance content (this utterance content is a kind of input information), but is not limited to the user's utterance content, and the keyboard. The input information may be input from the above. Therefore, the “utterance content” shown below can be described by replacing “utterance content” with “input information”.
[0054]
Similarly, in the following description, for convenience of explanation, the description will be limited to the “spoken sentence type” (speech type), but is not limited to this “spoken sentence type”, and input from a keyboard or the like. It may be an “input type” indicating the type of input information. Accordingly, the following “speech sentence type” (speech type) can be described by replacing “speech type” with “input type”.
[0055]
The input unit 100 is an acquisition unit that acquires input information from a user. In the present embodiment, a microphone, a keyboard, and the like are used. The input unit 100 is also a character recognition unit that specifies a character string indicating input information based on input information (other than voice) input from a user.
[0056]
Here, the input information means characters, symbols, voices and the like input through a keyboard or the like. Specifically, the input unit 100 specifies a character string indicating input information based on the input information (other than voice) input, and outputs the specified character string to the conversation control unit 300 as a character string signal. Further, the input unit 100 that has acquired the utterance content from the user (the utterance content is composed of voice and is a kind of input information) with a microphone or the like, the voice constituting the acquired utterance content is a voice signal. To the voice recognition unit 200.
[0057]
The voice recognition unit 200 is a character recognition unit that identifies a character string corresponding to the utterance content based on the utterance content acquired by the input unit 100. Specifically, the speech recognition unit 200 to which a speech signal is input from the input unit 100 analyzes the input speech signal, and a character string corresponding to the analyzed speech signal is stored in the speech recognition dictionary storage unit 700. The specified character string is output to the conversation control unit 300 as a character string signal. The speech recognition dictionary storage unit 700 stores a dictionary corresponding to standard speech signals.
[0058]
The sentence analysis unit 400 analyzes a character string specified by the input unit 100 or the speech recognition unit 200. In the present embodiment, as shown in FIG. 2, a morpheme extraction unit 410 and a phrase analysis unit 420 are included. A sentence structure analysis unit 430, an utterance type determination unit 440, a morpheme database 450, and an utterance type database 460.
[0059]
The morpheme extraction unit 410 is a morpheme extraction unit that extracts each morpheme constituting a minimum unit of a character string as first morpheme information based on the character string specified by the input unit 100 or the speech recognition unit 200.
[0060]
Specifically, the morpheme extraction unit 410 to which the character string signal is input from the management unit 310 extracts each morpheme from the character string corresponding to the input character string signal. Here, in this embodiment, the morpheme means the minimum unit of the word structure represented in the character string. As the minimum unit of the word structure, as shown in FIG. 3, for example, parts of speech such as nouns, adjectives and verbs can be cited. In the present embodiment, each morpheme is expressed as m1, m2,.
[0061]
That is, the morpheme extraction unit 410 collates a character string corresponding to the input character string signal with a morpheme group such as nouns, adjectives, and verbs stored in advance in the morpheme database 450, and the morpheme is extracted from the character string. Each morpheme (m1, m2,...) That matches the group is extracted, and each extracted morpheme is output to the phrase analysis unit 420 as an extraction signal.
[0062]
The phrase analysis unit 420 is a conversion unit that converts each morpheme into a phrase format based on each morpheme extracted by the morpheme extraction unit 410. Specifically, the phrase analysis unit 420 to which the extraction signal is input from the morpheme extraction unit 410 uses the morphemes corresponding to the input extraction signal to combine them into a phrase format.
[0063]
Here, in this embodiment, the phrase format is a sentence in which the independent grammar or one or more attached words are attached to the independent grammar in the Japanese grammar, or a character string that does not destroy the meaning of the Japanese grammar. Means a sentence that is separated as finely as possible. In this embodiment, this clause is expressed as p1, p2,..., Pi,.
[0064]
That is, the phrase analysis unit 420, as shown in FIG. 4, based on each morpheme corresponding to the input extracted signal, the dependency element of each morpheme (for example, (m2), (m4), (m5) ) ・ ・） Is extracted, and each morpheme is grouped into each phrase based on the extracted dependency elements. “T” shown in the figure means transposition.
[0065]
The phrase analysis unit 420 that collects each morpheme into each phrase includes a sentence structure analysis unit 430 and an utterance type determination unit 440 using sentence pattern information including each phrase that combines each morpheme and each morpheme constituting each phrase as a sentence pattern signal. Output to.
[0066]
The sentence structure analysis unit 430 is a classification unit that classifies each morpheme of the first morpheme information segmented by the phrase analysis unit 420 into attributes such as subject case and target case. Specifically, the sentence structure analysis unit 430, to which the sentence pattern signal is input from the phrase analysis unit 420, determines each morpheme included in the phrase based on each morpheme corresponding to the input sentence pattern signal and the phrase composed of each morpheme. Determine the “case composition”.
[0067]
Here, the “case structure” means a case (attribute) indicating a substantial concept in the clause. In the present embodiment, for example, a subject (subject) that represents a subject / subject, Examples include a meaning object (target case), an action meaning an action, a time meaning (tense, aspect), a location meaning a place, and the like. In this embodiment, each morpheme associated with the “case” (case configuration) of the three elements of the subject, the object, and the action is used as the first morpheme information.
[0068]
That is, as shown in FIG. 5, for example, when the dependency element of each morpheme is “” or “is”, the sentence structure analysis unit 430 applies the subject (subject) to the morpheme before the dependency element. Or a leading character). In addition, the sentence structure analysis unit 430
For example, if the dependency element of each morpheme is “NO” or “NO”, it is determined that the morpheme before the dependency element is an object (target).
[0069]
Further, for example, when the dependency element of each morpheme is “Yes”, the sentence structure analysis unit 430 includes an action (predicate; this predicate is composed of a verb, an adjective, and the like). It is determined that
[0070]
The sentence structure analysis unit 430 that has determined the “case structure” of each morpheme that constitutes each clause determines the range of topics (topics) to be described later based on the first morpheme information associated with the determined “case structure”. A topic search command signal for specifying is output to the reflective determination unit 320.
[0071]
The utterance type determination unit 440 is a type specifying unit that specifies an utterance type (input type) indicating the type of utterance content (input information) based on the phrase specified by the phrase analysis unit 420. Specifically, “speech sentence type” (speech type) is determined based on each morpheme corresponding to the sentence pattern signal input from the phrase analysis unit 420 and a phrase composed of each morpheme.
[0072]
In this embodiment, as shown in FIG. 6, the “spoken sentence type” is a statement sentence (D; Declaration), an impression sentence (I; Impression), a conditional sentence (C; Condition), and a result sentence ( E; Effect, time sentence (T; Time), location sentence (L; Location), repulsive sentence (N; Negation), and the like.
[0073]
The statement sentence means a sentence composed of a user's opinion or idea, and in this embodiment, as shown in FIG. 6, for example, a sentence such as “I like Sato” can be cited. An impression sentence means the sentence which consists of an impression which a user holds. A place sentence means a sentence made up of place elements.
[0074]
A result sentence means a sentence composed of sentences including a result element for a topic. A time sentence means a sentence composed of sentences including temporal elements related to a topic.
[0075]
The conditional sentence means a sentence composed of sentences including elements such as a premise of a topic, a condition and a reason why the topic is established, when one utterance is regarded as a topic. The repulsive sentence means a sentence composed of a sentence including an element that repels the utterance partner. An example sentence for each “spoken sentence type” is as shown in FIG.
[0076]
That is, the utterance type determination unit 440 collates each phrase with each dictionary stored in the utterance type database 460 based on each phrase corresponding to the input sentence pattern signal, and from each phrase, Extract sentence elements related to the dictionary. The utterance type determination unit 440 that extracts sentence elements related to each dictionary from each phrase determines “spoken sentence type” based on the extracted sentence elements. The sentence element means a type for identifying the type of character string, and the sentence element in the present embodiment includes the above-described definition phrase (which means “-”).
[0077]
Here, as shown in FIG. 7, the utterance type database 460 includes a definition expression example dictionary having a dictionary related to a definition phrase (for example,), an affirmative phrase (for example, approval, sympathy, ping pong). positive cases dictionary with a dictionary related to, a result clause (for example, so, so) the results representation case dictionary with a dictionary related to, greeting clause (for example, Hello) greeting case dictionary with a dictionary related to, It is composed of a negative example dictionary having a dictionary related to a negative phrase (for example, it is not idiot, the opposite), and each dictionary is associated with “type of spoken sentence”.
[0078]
As a result, the utterance type determination unit 440 compares the phrase with each dictionary stored in the utterance type database 460, extracts the sentence element related to each dictionary from the phrase, and associates it with the extracted sentence element. The “spoken sentence type” can be determined by referring to the determined type.
[0079]
The utterance type determination unit 440 outputs an answer search command signal for causing the corresponding user to search for a specific answer sentence to the answer sentence search unit 370 based on an instruction from the topic search unit 360 described later.
[0080]
The conversation database 500 stores a plurality of second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof, and a plurality of reply contents to the user with respect to utterance contents in association with each other in advance. Means (discourse storage means). The conversation database 500 is also an answer storage unit (discourse storage unit) that stores in advance a plurality of answer types indicating the types of answer contents associated with a plurality of answer contents in association with the second morpheme information.
[0081]
Furthermore, the conversation database 500 stores in advance a plurality of discourse ranges (keywords) indicating morphemes that constitute a range relevant to the input content that will be input by the user or the response content to the user. But there is. A plurality of second morpheme information indicating a morpheme made up of one character, a plurality of character strings, or a combination thereof is associated with the discourse range (keyword), and each second morpheme information includes a response content to the user. Associated with each.
[0082]
Furthermore, the conversation database 500 is also an answer storage means (discourse storage means) for classifying and storing each element constituting the second morpheme information into attributes such as each subject consisting of a main case and a target case consisting of a target case. .
[0083]
As shown in FIG. 8, in the present embodiment, the conversation database 500 is roughly divided in the present embodiment, which means a range that is related to the content of the speech that will be spoken by the user or the content of the response to the user. It consists of a range (discourse) and a topic (topic) that means a range that is most closely related to the content the user is speaking. As shown in the figure, “discourse range” is positioned as a superordinate concept of “topic” in the present embodiment.
[0084]
Each discourse range can be configured to have a hierarchical structure as shown in FIG. As shown in the figure, for example, a higher level discourse range (entertainment) for a certain discourse range (movie) is positioned in the upper hierarchical structure, and a lower level discourse range (movie) for the discourse range (movie). The attribute (movie) can be located in the lower hierarchical structure. That is, in the present embodiment, each discourse range can be arranged at a hierarchical position where the relationship between the higher concept, the lower concept, the synonym, and the synonym becomes clear with other discourse ranges.
[0085]
As described above, the discourse range is composed of each topic. In the present embodiment, for example, if the discourse range is an A movie name, it includes a plurality of topics related to “A movie name”. .
[0086]
This topic means a morpheme composed of a single character, a plurality of character strings, or a combination thereof, that is, a morpheme that constitutes the utterance content that will be uttered by the user. (Subject case), object (target case), and action “case” (attribute) are associated with each morpheme. In this embodiment, each morpheme associated with these three elements is expressed as a topic title (this topic title corresponds to a subordinate concept of “topic”) (second morpheme information). .
[0087]
The topic title is not limited to each morpheme associated with the above three elements, but other “cases”, that is, time (tense, aspect) meaning time, location meaning place, You may have each morpheme matched with the condition which means a condition, the impression which means an impression, the effect which means a result, etc.
[0088]
In this embodiment, the topic title (second morpheme information) is stored in advance in the conversation database 500, and the first morpheme information (derived from the utterance content uttered by the user) is It is a distinction.
[0089]
For example, if the talk range is “A movie name”, as shown in FIG. 10, the subject title is subject (A movie name), object (director), action (great) {this is “A movie name” The director of the name is composed of "meaning great".
[0090]
If there is no morpheme associated with “case composition” (subject, object, action, etc.) among the topic titles, “*” is indicated for the portion in the present embodiment.
[0091]
For example, {A What is a movie name? } Is converted into a topic title (subject; object; action). }, “A movie name” can be specified as a subject, but “object” and “action” are not elements of the sentence, so the topic title is “subject” (A movie name). “No object” (*); “No action” (*) (see FIG. 10).
[0092]
The answer sentence means an answer sentence (answer contents) to be answered to the user, and is associated with each topic title (second morpheme information) in this embodiment (see FIG. 8). In this embodiment, as shown in FIG. 11, the answer sentence is a statement sentence (D; Declaration) or an impression sentence (I; Impression) in order to make an answer corresponding to the type of utterance sentence uttered by the user. , Conditional sentence (C; Condition), result sentence (E; Effect), time sentence (T; Time), location sentence (L; Location), negative sentence (N; Negation), etc. ing.
[0093]
That is, as shown in FIG. 12, each answer sentence is associated with, for example, a discourse range (Sato) {subordinate concept; home run, superordinate concept; grass baseball, synonym; panda Sato, Sato player, panda} and each topic title. It has been.
[0094]
As shown in the figure, for example, the topic title 1-1 is {(Sato; *; I like): this consists of the order of (subject; object; action) as described above. If the order is the same below, the answer sentence 1-1 corresponding to the topic title 1-1 is (DA; statement affirmation “I like Sato”), (IA; comment affirmation) “I like Sato very much”), (CA; conditional affirmation “Sato ’s home run is very impressive”), (EA; a result affirming “I always watch Sato's games on TV”), (TA: Time affirmative sentence "I actually like it from the five-bats continual refrain in Koshien"), (LA; Place affirmative sentence "I like the serious face when standing on the blow"), (NA A repulsive affirmative sentence "I don't want to talk to people who don't like Sato, goodbye").
[0095]
In the present embodiment, as shown in FIG. 2, the conversation control unit 300 includes a management unit 310, a reflexive determination unit 320, a turnback determination unit 330, a discourse range determination unit 340, and an abbreviated sentence complement unit 350. , A topic search unit 360 and an answer sentence search unit 370.
[0096]
The management unit 310 controls the entire conversation control unit 300. Specifically, the management unit 310 to which the character string signal is input from the input unit 100 or the voice recognition unit 200 outputs the input character string signal to the morpheme extraction unit 410. In addition, the management unit 310 outputs the answer sentence searched by the answer sentence search unit 370 to the output unit 600.
[0097]
The reflexive determination unit 320 is a fixed form acquisition unit that collates the first morpheme information extracted by the morpheme extraction unit 410 with each fixed content, and searches the fixed content including the first morpheme information from each fixed content. .
[0098]
Here, the standard content means reflection element information for replying a standard content to the utterance content from the user, and this reflection element information is stored in the reflection element database 801 (standard storage means) in advance. A plurality are stored. As the reflective element information, in the present embodiment, as shown in FIG. 13, for example, "Good morning", "Hello", "Good evening", "greeting elements" such as "Hey", "I see", "Really?""Typicalelements" such as
[0099]
Specifically, the reflexive determination unit 320 to which the topic search command signal is input from the sentence structure analysis unit 430 receives the first morpheme information included in the input topic search command signal and the reflection element database 801. The reflection element information is collated, the reflection element information including the first morpheme information is searched from the reflection element information, and the searched reflection element information is output to the management unit 310.
[0100]
That is, assuming that the reflection element information is D1 and the first morpheme information is W, the reflective determination unit 320 determines that the relationship of W∩D1 ≠ φ (φ; empty set) is established. Process to give a typical answer.
[0101]
For example, when the user utters the utterance content “Good morning”, the reflexive determination unit 320 collates the utterance content “Good morning” with each reflection element information, and the utterance content from each reflection element information. The reflection element information “good morning” including (matching) “good morning” is searched, and the searched reflection element information “good morning” is output to the management unit 310.
[0102]
When the reflection determination unit 320 cannot search the reflection element information including the utterance content from the reflection element information, the reflex determination unit 320 returns the topic search command signal input from the sentence structure analysis unit 430. Output to.
[0103]
The return determination unit 330 collates the current first morpheme information extracted by the morpheme extraction unit 410 with the past response content stored in the return element database 802, and the current first morpheme information is a past response. If included in the content, it is a fixed form acquisition means for acquiring the content of the agreement.
[0104]
Here, “turnback” means to say back the content of the user's utterance as it is (or content close to it) in this embodiment. In this embodiment, the return element is composed of the first morpheme information that constitutes the response content output from the conversation control device 1 immediately before. For example, as shown in FIG. (Horse; *; beautiful), “I like Sato” (Sato; *; I like).
[0105]
Further, the return element database 802 is also an agreement storage means for preliminarily storing agreement contents for agreeing to input information input by the user. The content of the agreement includes, for example, the input information previously input by the user (if the previous input information by the user is “A film name director is Mr. S”, the content of the agreement is “ The director of the A movie name is Mr. S)), or “Yes”, “It is true”.
[0106]
Specifically, the return determination unit 330 to which the topic search command signal is input from the reflective determination unit 320 configures the first morpheme information and the return element included in the input topic search command signal for each return element. Each morpheme is collated, and it is determined whether or not the first morpheme information is included in the return element (see FIG. 14).
[0107]
If it is determined that the first morpheme information is included in each return element, the return determination unit 330 acquires the content of the agreement and outputs an answer sentence including the acquired content of the agreement to the management unit 310 (return) Process). That is, assuming that the return element (such as the previous answer sentence) is S and the first morpheme information is W, the return determination unit 330 indicates the above when the relationship of W⊂S and S ≠ φ is established. Performs a reversal process.
[0108]
For example, the conversation control device 1 responds with “The director of the A movie name is Mr. S” (Director of the A movie name; Mr. S; *) (the order is the subject, the object, the order of the action, and so on. ) Is output, and then the user responds with a response if the utterance is “is the director of A movie name Mr. S” (director of A movie name; Mr. S; *)? Since the determination unit 330 matches the first morpheme information of the user (Director of A movie name; Mr. S; *) and each morpheme of the answer sentence (Director of A movie name; Mr. S; *), The user concludes that he / she is replying to the contents of the reply, obtains the stored agreement contents “Yes”, and outputs the obtained agreement contents.
[0109]
Further, the return determination unit 330 collates the current first morpheme information extracted by the morpheme extraction unit 410 with the past first morpheme information stored in the return element database 802 to obtain the current first morpheme information. Is included in the past first morpheme information, it is also a fixed form acquisition means for acquiring the repulsion content.
[0110]
Specifically, when the user utters the utterance content “Horse is beautiful” and the conversation control device 1 outputs the content “Horse is good” as the response content When the person repeats the utterance content that “the horse is beautiful”, the turn-back determination unit 330 determines that each morpheme (first morpheme information) {horse; *; beautiful} constituting the current utterance content “horse is beautiful” Each morpheme (first morpheme information) {horse; *; beautiful} that constitutes the content of the previous utterance “Horse is beautiful” matches the content of the response from the conversation control device 1 to the user. You can conclude that you have not heard at all about "I feel good."
[0111]
In this case, since the user has not heard the response content from the conversation control device 1, the return determination unit 330 acquires the stored repulsion content (for example, do not repeat the same content), and the acquired repulsion The contents can be output.
[0112]
On the other hand, if the reversion determining unit 330 determines that the first morpheme information is the same as the content of the previous answer sentence or the first morpheme information is not the same as the previous first morpheme information, the reflex determination unit 320 The input topic search command signal is output to the conversation range determination unit 340.
[0113]
In addition, although the said return determination part 330 has shown the process in case a user makes a return with respect to "the reply content of the conversation control apparatus 1," the following processes can also be performed. For example, when the output unit 600 outputs an answer sentence “Horse is beautiful”, the user answers “Why is the horse beautiful?”, “Why is it beautiful?”, Or “Why?” This is a process of the turn-back determination unit 330 performed when the utterance is made.
[0114]
In this case, the return determination unit 330 outputs the response sentence S “the horse is beautiful” and the utterance content W from the user (“Why is the horse beautiful? (Question sentence)” or “Why is it beautiful? (Question sentence) ”), (W−c) ⊂S, S ≠ φ, c ≠ φ (where c means the utterance type of W, and this utterance type is determined by the utterance type determination unit 440 described later. As will be described later, for example, a question sentence can be cited as an utterance type, so a “conditional” reversal processing (the user asks the question contents for the answer contents). (Processing when a flip is performed).
[0115]
For example, when the conversation control device 1 outputs an answer sentence “The horse is beautiful”, the above-mentioned user utters the content of the utterance “Why is the horse beautiful?”. At that time, in order to resolve the user's question, the return determination unit 330 acquires an answer sentence such as “Even the horse is not beautiful” from the return element database 802, and the acquired answer sentence is sent to the management unit 310. Process to output.
[0116]
The discourse range determination unit 340 is a discourse search unit that collates the first morpheme extracted by the phrase analysis unit 420 and each discourse range, and searches the discourse range including the first morpheme information from each discourse range. .
[0117]
Specifically, the conversation range determination unit 340 to which the topic search command signal is input from the turn-back determination unit 330 determines the user's conversation range based on the input talk search command signal. That is, based on the input search command signal, the conversation range determination unit 340 searches the conversation database 500 for a range (discourse range) relevant to the content spoken by the user.
[0118]
For example, the discourse range determination unit 340 includes the first morpheme information included in the input topic search command signal (an interesting movie; *; is there) {is there an interesting movie? }, The first morpheme information is compared with the discourse range group, and if the discourse range group includes a morpheme (for example, “movie”) constituting the first morpheme information, The “movie” included in the information is determined as the discourse range. In this case, since the discourse range “movie” is included in the first morpheme information, the discourse range determination unit 340 includes the input first morpheme information in the topic search command signal and outputs it to the topic search unit 360.
[0119]
On the other hand, when the first morpheme information does not include a discourse range group, the conversation range determination unit 340 includes the input first morpheme information in the topic search command signal and outputs it to the abbreviated sentence complement unit 350.
[0120]
Thereby, the topic search unit 360 described later collates each “topic title” belonging to the “discourse range” determined by the conversation range determination unit 340 with the first morpheme information specified by the sentence structure analysis unit 430. Therefore, it is unnecessary to collate “all” “topic titles” (second morpheme information) with the first morpheme information, and the answer sentence search unit 370 to be described later will search for the final answer sentence. Can be shortened.
[0121]
As described above, the discourse range determination unit 340 collates the first morpheme information and the discourse range group, and if the discourse range group includes the morpheme of the first morpheme information, determines the morpheme as the discourse range. However, the present invention is not limited to this, and the morpheme of the return element searched immediately by the return determination unit 330 or the morpheme constituting the utterance content spoken by the user immediately before is determined as the discourse range. Also good. The abbreviated sentence complementing unit 350, which will be described later, can add the discourse range to the first morpheme information from which the morpheme is omitted, using the discourse range determined by the discourse range determination unit 340.
[0122]
The abbreviated sentence complementing unit 350 searches the attributes (subjects, objects, actions, etc.) constituting the first morpheme information based on the first morpheme information extracted by the phrase analysis unit 420 for attributes that do not include a morpheme. Attribute search means. The abbreviated sentence complementing unit 350 is also a morpheme adding unit that adds, to the attribute, a morpheme that constitutes the discourse range previously searched by the discourse range determining unit 340 based on the searched attribute.
[0123]
Specifically, the abbreviated sentence complementing unit 350 to which the topic search command signal is input from the discourse range determination unit 340 is based on the first morpheme information included in the input discourse search command signal, and the utterance including the first morpheme information. It is determined whether or not the content is an abbreviated sentence. If the utterance content including the first morpheme information is an abbreviated sentence, the morpheme in the discourse range to which the first morpheme information belongs is added to the first morpheme information.
[0124]
For example, the abbreviation sentence complementing unit 350 has a morpheme constituting the first morpheme information included in the input topic search command signal (director; *; *) (director?) (This sentence is “what” director If it is unknown, it means an abbreviated sentence.), The discourse range previously determined by the discourse range determination unit 340 (A movie name; this A movie name indicates the title of the movie) If the first morpheme information belongs to the first morpheme information, the determined discourse range (A movie name) is added to the first morpheme information (director of “A movie name”); *; *)
[0125]
That is, assuming that the first morpheme information is W and the determined discourse range is D, the abbreviated sentence complementing unit 350 adds the discourse range D to the first morpheme information W, and uses the first morpheme information after the addition as a topic search command. It is included in the signal and output to the topic search unit 360.
[0126]
Thereby, even if the first morpheme information is an abbreviated sentence and it is not clear as Japanese, the abbreviated sentence complementing unit 350, when the first morpheme information belongs to a certain discourse range, D (A movie name) is added to the first morpheme information W (director; *; *), and the first morpheme information is W ′ (director of A movie name; *; *) {What is the director of A movie name? }, Even if the user's utterance content is an abbreviated sentence, the abbreviated sentence can be supplemented based on the previously determined discourse range, and the abbreviated sentence should be clarified Can do.
[0127]
For this reason, the first morpheme information is set so that the abbreviated sentence complementing unit 350 has the proper utterance contents constituting the first morpheme information even if the utterance contents constituting the first morpheme information are abbreviated sentences. Since the specific morpheme can be complemented, the topic search unit 360 acquires the optimal “topic title” (second morpheme information) related to the first morpheme information based on the first morpheme information after the complement. The answer sentence search unit 370 can output the answer content more suitable for the user's utterance content based on the “topic title” acquired by the topic search unit 360.
[0128]
The topic search unit 360 collates the first morpheme information extracted by the phrase analysis unit 420 or the first morpheme information supplemented by the abbreviated sentence complement unit 350 with each second morpheme information, It is the 1st search means which searches the 2nd morpheme information containing the morpheme which comprises 1st morpheme information from the inside.
[0129]
Specifically, the topic search unit 360 to which the topic search command signal is input from the talk range determination unit 340 or the abbreviated sentence complement unit 350 is based on the first morpheme information included in the input topic search command signal. The “topic title” including the morpheme of the first morpheme information is searched from each “topic title” (second morpheme information) belonging to the discourse range determined by the determination unit 340, and the search result is used as a search result signal. The information is output to the answer sentence search unit 370 and the utterance type determination unit 440.
[0130]
For example, when the “case configuration” constituting the first morpheme information is (Sato; *; I like) {I like Sato}, the topic search unit 360, as shown in FIG. Each morpheme (Sato; *; I like) belonging to the “composition” is compared with each topic title 1-1 to 1-4 belonging to the discourse range (Sato), and from among the topic titles 1-1 to 1-4 A topic title 1-1 (Sato; *; likes) that matches (or approximates) each morpheme (Sato; *; likes) belonging to the “case composition” is searched, and this search result is used as a search result signal as a response sentence. The data is output to the search unit 370 and the utterance type determination unit 440.
[0131]
The utterance type determination unit 440 to which the search result signal is input from the topic search unit 360, based on the input search result signal, an answer search command for searching for a specific answer sentence that answers to the corresponding user. A signal (this answer search command signal also includes the determined “sentence type”) is output to the answer sentence search unit 370.
[0132]
The answer sentence search unit 370 is an answer acquisition unit that acquires an answer sentence associated with the second morpheme information based on the second morpheme information (topic title) searched by the topic search unit 360. In addition, the answer sentence search unit 370 collates the utterance type of the identified user with each answer type associated with the second morpheme information based on the second morpheme information searched by the topic search unit 360, It is also a second search means for searching for an answer type that matches the user's utterance type from among the answer types.
[0133]
Specifically, the answer sentence search unit 370, to which the search result signal from the topic search unit 360 and the answer search command signal from the utterance type determination unit 440 are input, the topic title (search result) corresponding to the input search result signal. Based on the second morpheme information) and the “speech sentence type” (utterance type) corresponding to the answer search command signal, the answer sentence group (each answer content) associated with the “topic title”. Among them, an answer sentence having an answer type (this answer type means “type of answer sentence” shown in FIG. 11) that matches the “sentence sentence type” (DA, IA, CA, etc.) is searched.
[0134]
For example, when the topic title corresponding to the search result is the topic title 1-1 (Sato; *; I like) shown in FIG. 12, the answer sentence search unit 370 is associated with the topic title 1-1. Of the reply sentences 1-1 (DA, IA, CA, etc.), the reply composed of the reply type (DA) that matches the “spoken sentence type” (for example, DA; utterance type) determined by the utterance type determination unit 440. The sentence 1-1 (DA; (I also like Sato) is searched), and the searched answer sentence is output to the management unit 310 as an answer sentence signal.
[0135]
The management unit 310 to which the answer sentence signal is input from the answer sentence search unit 370 outputs the input answer sentence signal to the output unit 600. In addition, the management unit 310 to which the reflection element information is input from the reflective determination unit 320 or the contents of the turn-back process from the turn-back determination unit 330, the answer sentence corresponding to the input reflection element information, the content of the turn-back process that is input To the output unit 600.
[0136]
The output unit 600 is an output unit that outputs the answer text acquired by the answer text search unit 370. In this embodiment, for example, a speaker, a display, and the like can be given. Specifically, the output unit 600 to which an answer sentence is input from the management unit 310 outputs the input answer sentence {for example, I also like Sato}.
[0137]
(Conversation control method using conversation control device)
The conversation control method by the conversation control apparatus 1 having the above configuration can be implemented by the following procedure. FIG. 15 is a flowchart showing the procedure of the conversation control method according to the present embodiment.
[0138]
First, the input unit 100 performs a step of acquiring the utterance content from the user (S101). Specifically, the input unit 100 acquires the voice that constitutes the utterance content of the user, and outputs the acquired voice to the voice recognition unit 200 as a voice signal. Further, the input unit 100 identifies a character string corresponding to the input information (other than the voice) based on the input information (other than the voice) input from the user, and the conversation control unit uses the identified character string as a character string signal. Output to 300.
[0139]
Next, the voice recognition unit 200 performs a step of specifying a character string corresponding to the utterance content based on the utterance content acquired by the input unit 100 (S102). Specifically, the speech recognition unit 200 to which a speech signal is input from the input unit 100 analyzes the input speech signal, and a character string corresponding to the analyzed speech signal is stored in the speech recognition dictionary storage unit 700. The specified character string is output to the conversation control unit 300 as a character string signal.
[0140]
And the morpheme extraction part 410 performs the step which extracts each morpheme which comprises the minimum unit of a character string as 1st morpheme information based on the character string specified by the speech recognition part 200 (S103).
[0141]
Specifically, the morpheme extraction unit 410 to which the character string signal is input from the management unit 310 includes a character string corresponding to the input character string signal and nouns, adjectives, verbs, and the like stored in advance in the morpheme database 450. The morpheme group is collated, each morpheme (m1, m2,...) Matching the morpheme group is extracted from the character string, and each extracted morpheme is output to the phrase analysis unit 420 as an extraction signal.
[0142]
Then, the phrase analysis unit 420 performs a step of grouping each morpheme into a phrase format based on each morpheme extracted by the morpheme extraction unit 410 (S104). Specifically, the phrase analysis unit 420 to which the extraction signal is input from the morpheme extraction unit 410 uses the morphemes corresponding to the input extraction signal to combine them into a phrase format.
[0143]
That is, as shown in FIG. 4, the phrase analysis unit 420 extracts and extracts the dependency elements (for example, gahahahahaha ...) of each morpheme based on each morpheme corresponding to the input extraction signal. Each morpheme is grouped into each clause based on the dependency element. The phrase analysis unit 420 that collects each morpheme into each phrase includes a sentence structure analysis unit 430 and an utterance type determination unit 440 using sentence pattern information including each phrase that combines each morpheme and each morpheme constituting each phrase as a sentence pattern signal. Output to.
[0144]
Thereafter, the sentence structure analysis unit 430 performs a step of classifying each morpheme of the first morpheme information segmented by the phrase analysis unit 420 into attributes such as subject case and target case (S105). Specifically, the sentence structure analysis unit 430, to which the sentence pattern signal is input from the phrase analysis unit 420, determines each morpheme included in the phrase based on each morpheme corresponding to the input sentence pattern signal and the phrase composed of each morpheme. Determine the “case composition”.
[0145]
That is, as shown in FIG. 5, for example, when the dependency element of each morpheme is “” or “is”, the sentence structure analysis unit 430 applies the subject (subject) to the morpheme before the dependency element. Or a leading character). In addition, the sentence structure analysis unit 430
For example, if the dependency element of each morpheme is “NO” or “NO”, it is determined that the morpheme before the dependency element is an object (target).
[0146]
Further, for example, when the dependency element of each morpheme is “Yes”, the sentence structure analysis unit 430 includes an action (predicate; this predicate is composed of a verb, an adjective, and the like). It is determined that
[0147]
The sentence structure analysis unit 430 that has determined the “case structure” of each morpheme that constitutes each clause determines the range of topics (topics) to be described later based on the first morpheme information associated with the determined “case structure”. A topic search command signal for specifying is output to the topic search unit 360.
[0148]
Next, the utterance type determination unit 440 performs a step of specifying an utterance type indicating the type of utterance content based on the phrase specified by the phrase analysis unit 420 (S106). Specifically, “speech sentence type” (speech type) is determined based on each morpheme corresponding to the sentence pattern signal input from the phrase analysis unit 420 and a phrase composed of each morpheme.
[0149]
That is, the utterance type determination unit 440 collates each phrase with each dictionary stored in the utterance type database 460 based on each phrase corresponding to the input sentence pattern signal, and from each phrase, Extract sentence elements related to the dictionary. The utterance type determination unit 440 that extracts sentence elements related to each dictionary from each phrase determines “spoken sentence type” based on the extracted sentence elements.
[0150]
The utterance type determination unit 440 outputs an answer search command signal for causing the corresponding user to search for a specific answer sentence to the answer sentence search unit 370 based on an instruction from the topic search unit 360 described later.
[0151]
Next, the reflexive determination unit 320 compares the first morpheme information extracted by the morpheme extraction unit 410 with each fixed content, and performs a step of searching for fixed content including the first morpheme information from each fixed content. (S107: reflective processing).
[0152]
Specifically, the reflexive determination unit 320 to which the topic search command signal is input from the sentence structure analysis unit 430 receives the first morpheme information included in the input topic search command signal and the reflection element database 801. The reflection element information (standard content) is collated, the reflection element information including the first morpheme information is searched from each reflection element information, and the searched reflection element information is output to the management unit 310.
[0153]
When the reflection element information cannot be searched for the reflection element information including the first morpheme information from the reflection element information, the reflection determination unit 320 determines whether to return the topic search command signal input from the sentence structure analysis unit 430. To the unit 330.
[0154]
Next, the return determination unit 330 collates the first morpheme information extracted by the morpheme extraction unit 410 with each return element, and performs a step of searching for the return element including the first morpheme information from each return element ( S108;
[0155]
When it is determined that the first morpheme information is included in each return element, the return determination unit 330 acquires the return element including the first morpheme information, and manages the response sentence including the acquired return element. The data is output to the unit 310 (turnback processing). That is, if the return element (answer sentence output last time, utterance content spoken by the previous user, etc.) is S and the first morpheme information is W, the return determination unit 330 has a relationship of W⊂S and S ≠ φ. If it is established, the turn-up process shown above is performed.
[0156]
On the other hand, when determining that the first morpheme information is not included in each return element, the return determination unit 330 sends the topic search command signal input from the reflective determination unit 320 to the conversation range determination unit 340. Output.
[0157]
Then, the discourse range determination unit 340 compares the first morpheme extracted by the phrase analysis unit 420 with each discourse range, and performs a step of searching the discourse range including the first morpheme information from each discourse range. (S109).
[0158]
Specifically, the conversation range determination unit 340, to which the topic search command signal is input from the reversal determination unit 330, is based on the input search command signal, and the content the user is speaking from the conversation database 500. Search for a relevant range (discourse range).
[0159]
For example, the discourse range determination unit 340 includes the first morpheme information included in the input topic search command signal (an interesting movie; *; is there) {is there an interesting movie? }, The first morpheme information is compared with the discourse range group, and if the discourse range group includes a morpheme (for example, “movie”) constituting the first morpheme information, The “movie” included in the information is determined as the discourse range. In this case, since the discourse range “movie” is included in the first morpheme information, the discourse range determination unit 340 includes the input first morpheme information in the topic search command signal and outputs it to the topic search unit 360.
[0160]
On the other hand, when the first morpheme information does not include a discourse range group, the conversation range determination unit 340 includes the input first morpheme information in the topic search command signal and outputs it to the abbreviated sentence complement unit 350.
[0161]
Next, the abbreviation sentence complementing unit 350 includes attributes that do not include a morpheme among the attributes (subject, object, action, etc.) constituting the first morpheme information based on the first morpheme information extracted by the phrase analysis unit 420. Perform the step of searching. Thereafter, the abbreviated sentence complementing unit 350 performs a step of adding the morpheme constituting the discourse range searched by the discourse range determining unit 340 to the attribute based on the attribute not including the searched morpheme (S110; abbreviated sentence). Complement).
[0162]
Specifically, the abbreviated sentence complementing unit 350 to which the topic search command signal is input from the discourse range determination unit 340 is based on the first morpheme information included in the input discourse search command signal, and the utterance including the first morpheme information. It is determined whether or not the content is an abbreviated sentence. If the utterance content including the first morpheme information is an abbreviated sentence, the morpheme in the discourse range to which the first morpheme information belongs is added to the first morpheme information.
[0163]
For example, the abbreviation sentence complementing unit 350 has a morpheme constituting the first morpheme information included in the input topic search command signal (director; *; *) (director?) (This sentence is “what” director If it is unknown, it means an abbreviated sentence.), The discourse range determined by the discourse range determination unit 340 (A movie name; this A movie name indicates the title of the movie) If the first morpheme information belongs to a certain morpheme, the morpheme (A movie name) of the determined discourse range is added to the first morpheme information to the morpheme constituting the first morpheme information (director of “A movie name”); *; *)
[0164]
That is, assuming that the first morpheme information is W and the determined discourse range is D, the abbreviated sentence complementing unit 350 adds the discourse range D to the first morpheme information W, and uses the first morpheme information after the addition as a topic search command. It is included in the signal and output to the topic search unit 360.
[0165]
Next, the topic search unit 360 collates the first morpheme information extracted by the phrase analysis unit 420 or the first morpheme information supplemented by the abbreviated sentence complement unit 350 with each second morpheme information, and each second morpheme A step of searching for second morpheme information including morpheme constituting the first morpheme information from the information is performed (S111).
[0166]
Specifically, the topic search unit 360 to which the topic search command signal is input from the talk range determination unit 340 or the abbreviated sentence complement unit 350 is based on the first morpheme information included in the input topic search command signal. The “topic title” including the morpheme of the first morpheme information is searched from each “topic title” (second morpheme information) belonging to the discourse range determined by the determination unit 340, and the search result is used as a search result signal. The information is output to the answer sentence search unit 370 and the utterance type determination unit 440.
[0167]
For example, when the “case configuration” constituting the first morpheme information is (Sato; *; I like) {I like Sato}, the topic search unit 360, as shown in FIG. Each morpheme (Sato; *; I like) belonging to the “composition” is compared with each topic title 1-1 to 1-4 belonging to the discourse range (Sato), and from among the topic titles 1-1 to 1-4 A topic title 1-1 (Sato; *; likes) that matches (or approximates) each morpheme (Sato; *; likes) belonging to the “case composition” is searched, and this search result is used as a search result signal as a response sentence. The data is output to the search unit 370 and the utterance type determination unit 440.
[0168]
The utterance type determination unit 440 to which the search result signal is input from the topic search unit 360, based on the input search result signal, an answer search command for searching for a specific answer sentence that answers to the corresponding user. A signal (this answer search command signal also includes the determined “sentence type”) is output to the answer sentence search unit 370.
[0169]
Then, the answer sentence search unit 370 collates the utterance type of the identified user with each answer type associated with the second morpheme information based on the second morpheme information searched by the topic search unit 360, A step of searching for a response type that matches the utterance type of the user from each response type and obtaining an answer sentence associated with the searched response type is performed (S112).
[0170]
Specifically, the answer sentence search unit 370, to which the search result signal from the topic search unit 360 and the answer search command signal from the utterance type determination unit 440 are input, the topic title (search result) corresponding to the input search result signal. Based on the second morpheme information) and the “speech sentence type” (utterance type) corresponding to the answer search command signal, the answer sentence group (each answer content) associated with the “topic title”. Among them, an answer sentence having an answer type (this answer type means “type of answer sentence” shown in FIG. 11) that matches the “sentence sentence type” (DA, IA, CA, etc.) is searched.
[0171]
For example, when the topic title corresponding to the search result is the topic title 1-1 (Sato; *; I like) shown in FIG. 12, the answer sentence search unit 370 is associated with the topic title 1-1. Of the reply sentences 1-1 (DA, IA, CA, etc.), the reply composed of the reply type (DA) that matches the “spoken sentence type” (for example, DA; utterance type) determined by the utterance type determination unit 440. The sentence 1-1 (DA; (I also like Sato) is searched), and the searched answer sentence is output to the management unit 310 as an answer sentence signal.
[0172]
Next, the management unit 310 to which the answer sentence signal is input from the answer sentence search unit 370 outputs the input answer sentence signal to the output unit 600. In addition, the management unit 310 to which the reflection element information is input from the reflective determination unit 320 or the contents of the turn-back process from the turn-back determination unit 330, the answer sentence corresponding to the input reflection element information, the content of the turn-back process that is input To the output unit 600. The output unit 600 to which an answer sentence is input from the management unit 310 outputs the input answer sentence {for example, I also like Sato} (S113).
[0173]
(Operation and effect of conversation control system and conversation control method)
According to the invention according to the present application having the above-described configuration, the reflexive determination unit 320 collates the first morpheme information constituting the utterance content uttered by the user with each fixed content stored in advance, and each fixed content from among, it is possible to find the standard content including a first morpheme information, reflexive determination unit 320, for example when the first morpheme information is a fixed content such as "Hello" includes the standard content it is possible to answer the same boilerplate content "Hello" or the like.
[0174]
In addition, when the user's utterance content is a fixed content, the reflexive determination unit 320 replies with the fixed content (greeting etc.), so that the user first contacts the conversation control device 1. You can enjoy a sense of communication.
[0175]
Further, the reversal determination unit 330 compares the current first morpheme information with the past response content, and if the current first morpheme information is not included in the past response content, the agreement stored in advance is stored. Since the contents can be acquired, if the input information currently input from the user matches the past response contents, the return determination unit 330 returns the user to the previous response contents (use It is possible to conclude that the user has input the input information that the person is listening to the response content.
[0176]
In this case, the return determination unit 330 acquires the contents of the agreement that has been stored and the acquired agreement (for example, “That ’s right”) because the user has made a response to the previous response. Can be output. As a result, if the user does not understand the meaning of the answer content output from the conversation control device 1, he can listen again and listen again to the answer content, so he is also talking to other users. You can taste a feeling like this.
[0177]
Further, the reversal determination unit 330 collates the current first morpheme information with the past first morpheme information, and if the current first morpheme information is included in the past first morpheme information, the repulsion content is acquired. Therefore, when the input information input last time is included in the input information input this time, the user may repeatedly input the same content as the previous input information. Therefore, it can be determined that the user has not properly answered the response content from the conversation control device.
[0178]
In this case, since the user does not appropriately reply to the previous response content, the reversal determination unit 330 repels the user, so the stored repulsion content is acquired, and the acquired repulsion content Is output. As a result, if the user does not input appropriate input information with respect to the response content from the conversation control device 1, the repulsion content is output from the conversation control device 1. You can taste a sense of being.
[0179]
Further, in order to search for the second morpheme information approximate to the first morpheme information, the topic search unit 360 may collate each second morpheme information belonging to the “discourse range” with the first morpheme information. Since it is not necessary to collate the second morpheme information with the first morpheme information, the time until the second morpheme information approximate to the first morpheme information can be shortened.
[0180]
As a result, the topic search unit 360 can search (pinpoint search) the second morpheme information that approximates the first morpheme information in a short time, so that the answer sentence search unit 370 is the topic search unit 360. An answer sentence associated with the second morpheme information can be acquired in a short time based on the searched second morpheme information, and the conversation control device 1 quickly answers the utterance content from the user. be able to.
[0181]
Moreover, the topic search part 360 searches the 2nd morpheme information which contains the morpheme (element which comprises a user's utterance content) which comprises 1st morpheme information from each 2nd morpheme information, and an answer sentence search part Since 370 can acquire the answer content associated with the second morpheme information based on the second morpheme information searched by the topic search unit 360, the answer sentence search unit 370 can obtain the utterance content of the user. Based on each constituent morpheme (first morpheme information), it is possible to take into account the semantic spaces (subjects, objects, etc.) constructed by each morpheme and obtain the response contents created in advance based on such semantic spaces Thus, it is possible to acquire the answer content more suitable for the utterance content, rather than simply using the entire utterance content as a keyword and obtaining the answer content associated with the keyword.
[0182]
Further, since the topic search unit 360 searches for the second morpheme information including the first morpheme information, it is not necessary to search for the second morpheme information that completely matches the utterance content of the user, and the conversation control device 1 is developed. The developer who does not need to memorize a huge amount of answer contents corresponding to the utterance contents that will be uttered by the user in advance, and can reduce the capacity of the storage unit.
[0183]
Furthermore, the answer sentence search unit 370 selects the answer type that matches the user's utterance type from the answer types (description, affirmation, location, repulsion, etc.) associated with each second morpheme information belonging to the “discourse range”. Since the response content associated with the response type can be acquired based on the searched response type, the conversation control device 1 can determine the utterance type that constitutes the user's conversation content, for example, the user Answers that match the user's utterance type out of multiple answers based on what is simply stated, what the user has to say, what the user describes the location element, etc. As a result, it is possible to obtain an optimal answer to the corresponding user.
[0184]
Furthermore, the answer sentence search unit 370 matches the user's utterance type from the answer types associated with the second morpheme information belonging only to the “discourse range” searched by the discourse range determination unit 340. Since it is only necessary to search for the type (pinpoint search is possible), it is not necessary to search for the response type associated with the “all” second morpheme information and the user's utterance type one by one. Can be obtained in a short time.
[0185]
Finally, the abbreviated sentence complementing unit 350 belongs to the discourse range in which the first morpheme information exists even if the first morpheme information constituting the utterance content of the user is an abbreviated sentence and is not clear as Japanese. If so, the discourse range can be added to the first morpheme information, and the first morpheme information consisting of abbreviated sentences can be complemented.
[0186]
Thus, the abbreviated sentence complementing unit 350 allows the first morpheme information so that the utterance contents constituting the first morpheme information are appropriate Japanese even if the utterance contents constituting the first morpheme information are abbreviated sentences. Since a specific morpheme (such as a morpheme constituting a discourse range) can be complemented, the topic search unit 360 is based on the first morpheme information after completion supplemented by the abbreviated sentence complement unit 350. The optimal second morpheme information related to the information can be acquired, and the answer sentence search unit 370 can select the answer content more suitable for the user's utterance content based on the second morpheme information acquired by the topic search unit 360. Can be output.
[0187]
As a result, even if the input information from the user is an abbreviated sentence, the conversation control device 1 does not use a function such as a neuro network or an AI intelligence, and does not mean what the abbreviated sentence means through past search results. Since the developer of the conversation control device 1 does not need to install a neuro network or AI intelligence, the system of the conversation control device 1 can be constructed more simply.
[0188]
[Example of change]
In addition, this invention is not limited to the said embodiment, The change as shown below can be added.
[0189]
(First change example)
In this modification, the conversation database 500 may be an element storage unit that stores a plurality of pieces of element information indicating the entire set group including a plurality of morpheme sets in association with the set group. Further, the morpheme extraction unit 410 collates the morpheme extracted from the character string with each set group, selects a set group including the extracted morpheme from each set group, and selects an element associated with the selected set group. Information may be extracted as first morpheme information.
[0190]
As shown in FIG. 16, some morphemes included in the character string uttered by the user are similar. For example, as shown in FIG. 16, if the element information indicating the entire group is “gift”, the “gift” is similar to a present, gift, year-end gift, mid-year gift, celebration, etc. (group). Therefore, when there is a morpheme similar to “gift” (such as the present), the morpheme extraction unit 410 can handle the similar morpheme as “gift”.
[0191]
That is, for example, when the morpheme extracted from the character string is “present”, the morpheme extraction unit 410 has “present” as the element information representing “present” as shown in FIG. “Present” can be replaced with “Gift”.
[0192]
This allows the morpheme extraction unit 410 to sort out morphemes that are similar to each other, so that the developer who develops the conversation control device can perform the first corresponding to the semantic space that is grasped from the mutually similar first morpheme information. It is not necessary to create reply contents related to the bimorphic information and the second morpheme information one by one, and as a result, the amount of data stored in the storage unit can be reduced.
[0193]
(Second modified example)
As shown in FIG. 17, in the present modification example, the topic search unit 360 may include a ratio calculation unit 321 and a selection unit 322.
[0194]
The ratio calculation unit 321 compares the first morpheme information extracted by the morpheme extraction unit 410 with each second information, and the ratio of the first morpheme information to the second morpheme information for each second morpheme information. Is a calculation means for calculating.
[0195]
Specifically, the ratio calculation unit 321 to which the topic search command signal is input from the sentence structure analysis unit 430, based on the first morpheme information included in the input topic search command signal, as shown in FIG. One morpheme information is collated with each topic title (second morpheme information) belonging to the discourse range stored in the conversation database 500, and the first morpheme information occupies in each topic title for each topic title. Calculate the percentage.
[0196]
For example, as shown in FIG. 17, when the first morpheme information constituting the utterance sentence uttered by the user is (Sato; *; I like) {I like Sato}, the ratio calculation unit 321 Each morpheme belonging to “case composition” is matched with each morpheme belonging to “case composition” by comparing each morpheme belonging to “case composition” (Sato; *; likes) and each morpheme included in the topic title (Sato; *; likes). The ratio of (Sato; *; I like) is calculated to be 100%. The ratio calculation unit 321 performs these calculations for each topic title, and outputs each calculated ratio to the selection unit 322 as a ratio signal.
[0197]
The selection unit 322 is a selection unit that selects one second morpheme information from each second morpheme information according to the size of each rate calculated by the rate calculation unit 321 for each second morpheme information. is there.
[0198]
Specifically, the selection unit 322 to which the ratio signal is input from the ratio calculation unit 321 selects from among the ratios included in the input ratio signal (element of “case configuration” / element of “topic title” × 100). For example, a topic title with a high ratio is selected (see FIG. 18). The selection unit 322 that has selected a topic title with a high ratio outputs the selected topic title to the answer sentence search unit 370 and the utterance type determination unit 440 as a search result signal. The answer sentence search unit 370 acquires an answer sentence associated with the topic title based on the topic title selected by the selection unit 322.
[0199]
Thereby, the selection part 322 calculates the ratio which 1st morpheme information occupies with respect to 2nd morpheme information for every 2nd morpheme information, and sets it as the magnitude | size of each ratio calculated for each 2nd morpheme information. Accordingly, since one second morpheme information can be selected from each second morpheme information, the selection unit 322 has, for example, the first morpheme information (what constitutes the utterance content of the user) first. If the second morpheme information that accounts for a large proportion of the dimorphic information can be acquired from a plurality of second morpheme information groups, the second morpheme information that follows the semantic space grasped from the first morpheme information can be obtained more accurately. As a result, the answer sentence search unit 370 can make an optimum answer to the utterance content of the user.
[0200]
Moreover, since the selection part 322 can select a topic title with a high ratio calculated by the ratio calculation part 321 from a plurality of topic titles, it belongs to the “case configuration” included in the user's utterance sentence. Even if each morpheme and each topic title stored in the conversation database 500 do not completely match, it is possible to acquire a topic title closely related to each morpheme belonging to “case composition”.
[0201]
As a result, since the selection unit 322 can acquire a topic title closely related to the “case configuration” constituting the first morpheme information, the developer who develops the conversation control device 1 configures the first morpheme information. Since it is not necessary to store topic titles that completely match “case configuration” in the conversation database 500, the capacity of the conversation database 500 can be reduced.
[0202]
Furthermore, the ratio calculation unit 321 calculates the ratio of the first morpheme information to the second morpheme information for each second morpheme information belonging only to the “discourse range” searched by the discourse range determination unit 340. , It is not necessary to calculate the ratio of the first morpheme information to the “all” second morpheme information, and the second morpheme information following the semantic space composed of the first morpheme information is acquired in a shorter time. As a result, it is possible to quickly output the optimum answer contents for the utterance contents from the user based on the acquired second morpheme information.
[0203]
The ratio calculation unit 321 collates each morpheme of the first morpheme information belonging to each classified attribute and each morpheme of each second morpheme information belonging to each attribute stored in advance for each attribute, It may be a first search means for searching for second morpheme information including each morpheme of the first morpheme information in at least one attribute from the second morpheme information.
[0204]
Specifically, the ratio calculation unit 321 to which the topic search command signal is input, for each “case” (subject; object; action) of the “case configuration” included in the input topic search command signal. ”And each morpheme belonging to“ case ”of the topic title composed of the same“ case ”are collated, and it is determined whether or not the morphemes constituting each“ case ”are the same.
[0205]
For example, as illustrated in FIG. 19, when the “case” morpheme of “case configuration” is (dog; person; bitten) {dog bites a person}, the ratio calculation unit 321 includes those morphemes. Match "dog", "person", and "chewed" with the morpheme "dog", "person", and "chewed" of the topic title consisting of the same "case" that constitutes the morpheme Then, among the morphemes “dog”, “people”, and “chewing” that make up the topic title, the “case composition” morpheme “dog” consisting of the same “case” corresponding to each morpheme, The ratio of “people” and “biting” is calculated (100%).
[0206]
If the element constituting the topic title is (person; dog; bite) {person bites the dog}, the ratio calculation unit 321 performs morphemes belonging to the two cases by the same procedure as described above. Therefore, the degree of coincidence of each “case” between the morpheme constituting “case composition” and “topic title” is calculated to be 33% (see FIG. 19).
[0207]
The ratio calculation unit 321 that has calculated the ratio selects a topic title with a high ratio from each ratio, and outputs the selected topic title to the answer sentence search unit 370 and the utterance type determination unit 440 as a search result signal.
[0208]
As a result, the ratio calculation unit 321 displays each morpheme of the first morpheme information belonging to each classified “case configuration” (subject case, target case, etc.) and a pre-stored topic title for each “case”. Since the second morpheme information including each morpheme of the first morpheme information in at least one “case” can be searched from the topic titles, the ratio calculation unit 321 is different from the normal word order. If the utterance content is composed of things, for example, “a person bites a dog”, the morpheme of the subject case is “person” and the morpheme of the subject case is “dog”. You can search for matching second morpheme information, and obtain the response contents {"really?" Or "I don't know the meaning"} associated with the second morpheme information (person; dog; bite) can do.
[0209]
In other words, since the ratio calculation unit 321 can identify utterance contents that are difficult to identify, for example, “a person bites a dog” and “a dog bites a person”, an optimum answer to the identified utterance contents, For the former, for example, “Really?”, For the latter, for example, “Are you okay?”
[0210]
Further, the ratio calculation unit 321 may search for the second morpheme information including the morpheme of the first morpheme information in at least one attribute from the respective second morpheme information belonging to the “discourse range”. From the second morpheme information, it is no longer necessary to acquire one second morpheme information, the second morpheme information that follows the semantic space composed of the first morpheme information can be acquired in a shorter time, As a result, the conversation control device 1 can quickly output the optimum answer content for the utterance content from the user based on the acquired second morpheme information.
[0211]
Note that the selection unit 362 may select one topic title from among the topic titles according to a predetermined priority order. This priority order means a priority for selection as a topic title. This priority order is predetermined by the developer at the development stage.
[0212]
(Third change example)
As shown in FIG. 20, the present modification is not limited to the above embodiment and each of the modifications described above, and the communication unit 800 in the conversation control devices 1 a and 1 b and the communication unit via the communication network 1000. A communication unit 900 for transmitting / receiving data to / from 800, conversation databases 500b to 500d connected to the communication unit 900, and servers 2a to 2c may be provided (conversation control system).
[0213]
Here, the communication network 1000 means a communication network that transmits and receives data. In the present embodiment, for example, the Internet is used.
[0214]
In this modified example, the conversation control devices 1a and 1b, the conversation databases 500b to 500d, and the servers 2a to 2c are limited for convenience. However, the present invention is not limited to this, and another conversation database is provided. Also good. The servers 2a to 2c store contents similar to those stored in the conversation databases 500a to 500d.
[0215]
Thereby, the conversation control unit 300 not only includes the conversation database 500a arranged inside the conversation control device 1a but also the other conversation control device 1b, the conversation databases 500b to 500d, and the server 2a to the communication database 1000. 2c can be referred to, for example, it is possible to search the conversation database 500a for the discourse range related to each morpheme (first morpheme information) belonging to “case structure” included in the topic search command signal. Even if it is not possible, by referring to the other conversation control device 1b, the conversation databases 500b to 500d, and the servers 2a to 2c, it is possible to search the discourse range related to the first morpheme information, It is possible to search for an answer sentence more suitable for the utterance sentence.
[0216]
(Fourth change example)
The sentence structure analysis unit 430 may store in the conversation database 500 each “case configuration” and each “case configuration” constituting the identified first morpheme information. The answer sentence search unit 370 may store each “case structure” and each “morpheme” associated with each “case structure” constituting the searched answer sentence in the conversation database 500.
[0217]
The conversation range determination unit 340 may store the searched conversation range in the conversation database 500. The topic search unit 360 may store the searched second morpheme information in the conversation database 500.
[0218]
The first morpheme information, the second morpheme information, each “case composition” that constitutes the first morpheme information or the second morpheme information, each morpheme associated with each “case composition”, and the retrieved answer sentence Each “case structure” to be configured, each morpheme associated with each “case structure”, and the searched discourse range can be stored in the conversation database 500 as history morpheme information in association with each other.
[0219]
The abbreviated sentence complementing unit 350 does not include a morpheme from each attribute (subject, object, action, etc .; case configuration) constituting the first morpheme information based on the first morpheme information extracted by the phrase analysis unit 420. An attribute may be searched, and history morpheme information stored in the conversation database 500 may be added to the attribute based on the searched attribute.
[0220]
Specifically, the abbreviated sentence complementing unit 350 to which the topic search command signal is input from the discourse range determination unit 340 is based on the first morpheme information included in the input discourse search command signal, and the utterance including the first morpheme information. It is determined whether the content is an abbreviated sentence. If the utterance content including the first morpheme information is an abbreviated sentence (for example, the subject, object, or action does not have a predetermined morpheme), the conversation database 500 stores The stored history form information is added to the first morpheme information.
[0221]
That is, assuming that the subject included in the history form information is S1, the object is O1, the action A1, the discourse range is D1, and the omitted first morpheme information is W, the supplemented first morpheme information W1 is S1ＳW , O1∪W, A1∪W, or D1∪W.
[0222]
The topic search unit 360 collates the first morpheme information W1 supplemented by the abbreviated sentence complement unit 350 and each second morpheme information, and from each “topic title” (second morpheme information), the first morpheme information The second morpheme information including W1 is searched, and the searched topic title is output to the answer sentence search unit 370 and the utterance type determination unit 440 as a search result signal.
[0223]
Thereby, even if the utterance content composed of the first morpheme information is an abbreviated sentence and is not clear as Japanese, the abbreviated sentence complementing unit 350 uses the history form information stored in the conversation database 500, Since the omitted morpheme of the first morpheme information can be complemented, the utterance content including the omitted first morpheme information can be clarified.
[0224]
For this reason, when the abbreviated sentence complementing part 350 is the abbreviated sentence which comprises 1st morpheme information, 1st morpheme information is set so that the utterance content which consists of 1st morpheme information becomes appropriate Japanese. Since the morpheme omitted in FIG. 6 can be supplemented, the topic search unit 360 is based on the first morpheme information supplemented with the morpheme, and the optimal “topic title” (second morpheme) related to the first morpheme information. Information) can be acquired, and the answer sentence search unit 370 can output the answer content more suitable for the user's utterance content based on the optimal “topic title” acquired by the topic search unit 360. .
[0225]
(Fifth change example)
The topic search unit 360 may include a deletion unit 361 and a discourse addition unit 362 as shown in FIG. Based on the searched second morpheme information, the deletion unit 361 collates the second morpheme information with the discourse range searched by the discourse range determination unit 340, and from among the morphemes constituting the second morpheme information, It is a deletion means for deleting a morpheme that matches the range.
[0226]
Specifically, the topic search unit 360 to which the topic search command signal is input from the abbreviated sentence complement unit 350 receives the first morpheme information included in the input topic search command signal and the discourse determined by the discourse range determination unit 340. Each second morpheme information belonging to the range is collated, and second morpheme information matching the first morpheme information is searched from each second morpheme information.
[0227]
Then, based on the searched second morpheme information, the deletion unit 361 collates the second morpheme information with the morphemes constituting the discourse range determined by the discourse range determination unit 340, and stores the second morpheme information. Then, the morpheme that matches the morpheme constituting the discourse range is deleted, and the second morpheme information from which the morpheme is deleted is output to the discourse adding unit 362 as a delete signal.
[0228]
That is, the deletion unit 361 removes (removes) the current discourse range D2 (this D2 is made of morphemes) determined by the discourse range determination unit 340 from each morpheme t1 constituting the second morpheme information. If the result is t2, t2 = t1-D2).
[0229]
The discourse adding unit 362 acquires another discourse range associated with the discourse range searched by the discourse range determining unit 340 based on the second morpheme information from which the morpheme has been deleted by the deletion unit 361, This is a discourse adding means for adding the morpheme constituting the discourse range to the second morpheme information.
[0230]
More specifically, if the current discourse range D2 is related to the answer sentence K1, and the discourse range is DK, other discourses related to the answer sentence K1 or the current discourse range D2 (those that have a sibling relationship). Since the range D3 can be expressed as D3 = D2∪DK, the second morpheme information W2 after adding the morpheme constituting the other discourse range D3 can be set to W2 = t2∪D3.
[0231]
For example, each morpheme t1 constituting the second morpheme information is (A movie name; *; interesting) {A movie name is interesting? }, And the current discourse range D2 determined by the discourse range determination unit 340 is (A movie name), the deletion unit 361 first starts from each morpheme t1 (A movie name; *; interesting). The conversation range D2 (A movie name) is deleted, and the deleted result is set to t2 (*; *; interesting) (t2 = t1-D2).
[0232]
When the other conversation range D3 related to the current conversation range D2 (A movie name) is “B movie name”, the second morpheme information after adding the morpheme constituting the other conversation range D3 W2 is t2∪D3, so (B movie name; *; funny) {B movie name is interesting? }.
[0233]
Thus, when the user's utterance content is “A movie name is interesting?”, The discourse adding unit 362 matches each morpheme (A movie name; *; interesting) constituting the user's utterance content. Second morpheme information (A movie name; *; interesting), other second morpheme information (B movie name; *; interesting) {B movie name is interesting? }, The answer sentence search unit 370 acquires the answer sentence (for example, “B movie name is interesting”) associated with the second morpheme information changed by the discourse adding part 362, The acquired answer text can be output.
[0234]
As a result, the answer sentence search unit 370 does not output an answer sentence for the user's utterance content, but is related to the utterance content based on the second morpheme information including the morpheme added by the discourse addition unit 362. Since an answer sentence can be output, the output unit 600 can output a more humane answer sentence based on the answer sentence searched by the answer sentence search unit 370.
[0235]
Note that the discourse adding unit 362 is not limited to adding the other discourse range to the second morpheme information from which the morpheme has been deleted, but to the second morpheme information from which the morpheme has been deleted. (Stored in the database 500) may be added.
[0236]
(Sixth change example)
When the topic search unit 360 cannot search the second morpheme information including the first morpheme information from the second morpheme information, the topic search unit 360 collates the first morpheme information with each answer content, When the answer content including the first morpheme information can be searched from among the first morpheme information, the first search unit may acquire the second morpheme information associated with the searched answer content.
[0237]
Specifically, the topic search unit 360, to which the topic search command signal is input from the abbreviated sentence complement unit 350, based on the first morpheme information included in the input topic search command signal, If the second morpheme information that matches the first morpheme information cannot be obtained from each second morpheme information by collating with the morpheme information, as shown in FIG. The answer sentence associated with the second morpheme information is collated.
[0238]
When the topic search unit 360 determines that the morpheme constituting the first morpheme information (the action or the morpheme associated with the action) is included in the answer sentence by this collation, the answer sentence includes Search for associated second morpheme information.
[0239]
Thereby, even if the topic search unit 360 cannot search the second morpheme information that matches the first morpheme information from each second morpheme information, Since the answer sentence including the morpheme (action or the morpheme associated with the action) that constitutes the user can be identified and the second morpheme information associated with the identified answer sentence can be searched, the user's utterance content 2nd morpheme information corresponding to the 1st morpheme information which constitutes can be searched appropriately.
[0240]
As a result, since the topic search unit 360 can search for the optimal second morpheme information corresponding to the first morpheme information, the answer sentence search unit 370 can search for the optimal second morpheme information searched by the topic search unit 360. Based on the above, it is possible to acquire appropriate answer contents for the user's utterance contents.
[0241]
[program]
The contents described in the conversation control system and the conversation control method can be realized by executing a dedicated program for using a predetermined program language in a general-purpose computer such as a personal computer.
[0242]
Here, as the programming language, in this embodiment, the topic requested by the user, the degree of emotion held for the user in a certain matter, or the type of statement, affirmative sentence, question sentence, repulsive sentence, etc. Languages that are hierarchically stored in a database in association with morphemes, for example, DKML (Discourse Knowledge Markup Language), XML (eXtensible Markup Language), C language, etc., developed by the inventors.
[0243]
That is, the conversation control device 1 stores data stored in each of the conversation databases 500a to 500d (stored information such as second morpheme information, fixed content, answer sentence, answer type, set group, discourse range, element information, etc.), etc. These parts can be constructed by DKML (Discourse Knowledge Markup Language) or the like, and a program for using the constructed storage information or the like can be executed.
[0244]
According to such a program according to the present embodiment, each morpheme constituting the utterance content of the user is identified, the semantic content grasped from each identified morpheme is analyzed, and associated with the analyzed semantic content. A conversation control device, a conversation control system, and a conversation control method that have the effect of being able to output the optimum answer contents corresponding to the user's utterance contents by outputting the answer contents prepared in advance. It can be easily realized by a general purpose computer.
[0245]
In addition, since the developer who develops the conversation control device 1 can hierarchically construct the second morpheme information and the like for searching the answer contents for the user's utterance contents using the language in the database. The conversation control device 1 can acquire the response content for the utterance content from the database through a hierarchical procedure based on the utterance content of the user.
[0246]
In other words, the conversation control device 1 has determined and determined the hierarchy of the user's utterance content (for example, whether the second morpheme information stored in the database is in a higher concept or a lower concept). Appropriate answer contents can be acquired from the answer contents accumulated in advance based on the hierarchy.
[0247]
For this reason, the conversation control device 1 does not collate the first morpheme information composed of the user's utterance content and the previously stored “all” second morpheme information one by one. Since the second morpheme information and the first morpheme information may be collated, the second morpheme information approximate to the first morpheme information can be acquired in a short time.
[0248]
Furthermore, communication between the communication unit 800 and the communication unit 900 may be performed by transmitting and receiving data via a communication network 1000 using a protocol such as DKML. Thereby, for example, when there is no answer content suitable for the user's utterance content in the conversation control device 1, the conversation control device 1 is suitable for the user's utterance content through the communication network 1000 according to the convention such as DKML. It is possible to retrieve the answer contents (described in DKML or the like) and obtain the retrieved answer contents (see FIG. 20).
[0249]
The program can be recorded on a recording medium. As shown in FIG. 23, examples of the recording medium include a hard disk 1100, a floppy disk 1200, a compact disk 1300, an IC chip 1400, a cassette tape 1500, and the like. According to the recording medium on which such a program is recorded, the program can be easily stored, transported, sold, and the like.
[0250]
[Second Embodiment]
(Basic configuration of ranked system)
A second embodiment of the present invention will be described with reference to the drawings. FIG. 24 shows the internal structure of the ranked system according to this embodiment. As shown in the figure, the ranking system is almost the same as the internal structure of the conversation control device 1 (ranking device) in the first embodiment, but the ranking unit 380 (ranking means) is added to the conversation control unit 300. It is different in having. Since it is the same as the structure of 1st embodiment and a modified example except this difference, description about structures other than a difference is abbreviate | omitted.
[0251]
In the first embodiment, the conversation control device 1 has been described with respect to the process of acquiring the optimum answer sentence corresponding to the input information based on the input information from the user. A description will be given of a process in which the ranking unit 380 ranks the emotion level and the like of the user based on the input information from the user. The specific explanation is as follows.
[0252]
The ranking unit 380 ranks the emotion level and the like held for the user based on the input information input from the user. In this embodiment, as shown in FIG. A unit 381, an understanding state determination unit 382, and a conflict state determination unit 383. Here, the emotion level means the level of emotion that the conversation control apparatus 1 has for the input information. In the present embodiment, this emotion level is uniquely determined by an emotion state determination unit 381 described later based on the input information.
[0253]
The ranking unit 380 determines the degree of emotion held by the user based on the “Discourse Event” table stored in the discourse database. Here, as shown in FIG. 25, the “discourse event” table includes a cohesion factor (cohere) indicating that the first morpheme information is a factor having relevance (cohesion) with the discourse range, as shown in FIG. It consists of a topic change (shift) indicating that the first morpheme information is a factor that is not related to the current discourse range.
[0254]
Specifically, as shown in FIG. 26, the binding factor includes a binding factor (−) and a binding factor (+) in the present embodiment. In the cohesive factor (−), as shown in the figure, the topic search unit 360 searches the topic title associated with “low emotional level” belonging to the first discourse range in which the first morpheme information belongs to the previous discourse range. Means that. Here, “low emotion level” means that the conversation control device 1 sets a low emotion level for the topic title (for example, when the conversation control device 1 is not very interested in the topic title). Means that
[0255]
In the cohesive factor (+), as shown in the figure, the topic search unit 360 searches for the topic title associated with the “high emotion level” belonging to the first discourse range in which the first morpheme information belongs. Means that. Here, “high emotion level” means that the conversation control device 1 sets a high emotion level for the topic title (for example, when the conversation control device 1 is interested in the topic title). Means that.
[0256]
As shown in FIG. 26, these binding factors (−) and (+) are ranked based on the result of the “conditional” return processing by the return determination unit 330 and the result of the abbreviated sentence complement processing by the abbreviated sentence complement unit 350. Part 380.
[0257]
Moreover, the topic change is made up of topic change (−) and topic change (+) as shown in FIG. The topic change (−) means that the current topic title searched by the topic search unit 360 is changed to another topic title associated with “low emotion level” later, as shown in FIG. .
[0258]
The topic change (+) means that the current topic title searched by the topic search unit 360 is later changed to another topic title associated with “high emotion level” as shown in FIG. . These topic changes (−) (+) are determined by the ranking unit 380 based on the search results of the talk range by the talk range determination unit 340, as shown in FIG.
[0259]
In addition to the above, the “Discourse Scope Event” table has an interrupt factor indicating that the first morpheme information interrupts the current topic, and the first morpheme information is more difficult than the second morpheme information. This is composed of a difficult factor (harass) indicating that the first morpheme information is a conflicting factor (oppose) indicating that the first morpheme information is a conflicting factor with respect to the second morpheme information.
[0260]
As illustrated in FIG. 24, the interruption factor is determined by the ranking unit 380 based on the turn-back process performed by the turn-up determination unit 330 and the reflective process performed by the reflective determination unit 320. The difficult factor is determined by the ranking unit 380 based on the search result of the topic title (second morpheme information) by the topic search unit 360. The conflict factor is determined by the ranking unit 380 based on the determination of the utterance type by the utterance type determination unit 440.
[0261]
The emotional state determination unit 381 is a ranking unit that ranks the emotion level held by the user based on the search result by the topic search unit 360. Here, the emotion level determined by the emotion state determination unit 381 can be ranked in six levels as shown in FIG. This emotion level is positioned in the direction in which the numerical value shown in the figure increases as the emotion level held by the user increases.
[0262]
Therefore, when the emotion level transitions from e (0) to e (3), it means that the emotion level held for the user increases. On the other hand, when the emotion level transitions from e (0) to e (-2), it means that the emotion level held for the user is lowered. Of the six levels, the reference emotion level is e (0) in the present embodiment.
[0263]
In the present embodiment, the ranking of the emotion level is performed using an emotion minus factor, an emotion plus factor, and topic change (+) (−). In this embodiment, this negative emotion factor is assumed to be the above-described interruption factor, difficult factor, conflict factor, cohesion factor (-), and topic change (-). The emotion plus factor consists of a cohesion factor (+) and a topic change (+).
[0264]
In the figure, the topic level (-) (+) changes the emotion level by two or more levels. This indicates that the current topic title is “Emotion level” is very “low” or “Emotion level” is very high. When a topic title associated with “high” is selected, the emotion level is changed in two or more stages.
[0265]
Specifically, when the second morpheme information including the first morpheme information is not searched by the topic search unit 360, the emotion state determination unit 381 ranks the emotion level held by the user to a lower rank.
[0266]
That is, the fact that the second morpheme information including the first morpheme information is not searched by the topic search unit 360 means that each of the second morpheme information in which the first morpheme information constituting the input information input from the user is stored in advance. Therefore, the emotion state determination unit 381 determines that the input information (first morpheme information) input from the user is an emotion negative factor (difficult factor). , Set the emotion level for the user to be lower.
[0267]
On the other hand, when the second morpheme information including the first morpheme information is searched by the topic search unit 360, the emotion state determination unit 381 ranks the emotion level held for the user in a high rank.
[0268]
That is, the fact that the second morpheme information including the first morpheme information is searched by the topic search unit 360 means that each second morpheme in which the first morpheme information constituting the input information input from the user is stored in advance is stored. Since it can be determined that the information is easier than the information, the emotion state determination unit 381 determines that the input information input from the user can be understood, and has a higher feeling of feeling for the user. Set to.
[0269]
The emotion state determination unit 381 is also a ranking unit that ranks the emotion level held for the user based on the determination result by the conversation range determination unit 340 (relevance determination unit). In the present embodiment, the discourse range determination unit 340 collates the first discourse range indicating the searched discourse range with the second discourse range indicating the searched discourse range, and the second discourse range is the first discourse range. Relevance determining means for determining whether or not there is an association with a range.
[0270]
Specifically, the emotion state determination unit 381 has a low feeling of feeling for the user when the conversation range determination unit 340 determines that the second conversation range is not related to the first conversation range. Rank to rank.
[0271]
That is, the fact that the second topic range is determined not to be related to the first topic range by the conversation range determination unit 340 means that the current conversation range (second conversation range) including the first morpheme information is the previous one. Can be determined to be different from the discourse range (the first discourse range) of (the contextual topic has no unity surname; this is also referred to as “topic change”). If the first morpheme information belonging to the second discourse range is included in the topic title (second morpheme information) associated with “low emotion level”, the first morpheme information is an emotion negative factor (topic change (−)) And the emotion level held for the user is set to be lower (see FIGS. 26 and 27).
[0272]
Thus, the topic change (-) indicates that the current conversation range is changed to another conversation range, and the first morpheme information belonging to the changed conversation range is content that is not interesting for the conversation control device 1. means.
[0273]
In this case, if the first morpheme information belonging to the second discourse range is included in the topic title (second morpheme information) associated with “high emotion level”, the emotion state determination unit 381 determines that the first morpheme information is It can be determined that it is an emotional plus factor (topic change (+)), and the emotional level held for the user can be set higher.
[0274]
As a result, topic change (+) means that the current discourse range is changed to another discourse range, and the first morpheme information belonging to the changed discourse range has interesting contents for the conversation control device 1. To do.
[0275]
On the other hand, the emotional state determination unit 381 ranks the emotion level held by the user in a high rank when the conversation range determination unit 340 determines that the second conversation range is related to the first conversation range. Attached.
[0276]
In other words, the fact that the second topic range is determined to be related to the first topic range by the conversation range determination unit 340 means that the current conversation range (second conversation range) including the first morpheme information is the previous one. Since it can be determined that it is substantially the same as the discourse range (first discourse range) (contextual topics have cohesiveness), the emotional state determination unit 381 has the second discourse range. When one morpheme information is included in the topic title (second morpheme information) associated with “high emotion level”, it is determined that the first morpheme information is an emotion plus factor (cohesion factor (+)) and used The emotion level held for the person is set higher (see FIGS. 26 and 27).
[0277]
Thus, the cohesion factor (+) is that the current conversation range is substantially the same as the previous conversation range, and the first morpheme information belonging to the current conversation range is interesting for the conversation control device 1. Means.
[0278]
In this case, when the first morpheme information belonging to the second discourse range is included in the topic title (second morpheme information) associated with “low emotion level”, the emotion state determination unit 381 determines that the first morpheme information is It can be determined that it is an emotion negative factor (cohesion factor (-)), and the emotion level held for the user can be set to be lower.
[0279]
As a result, the cohesion factor (−) is such that the current discourse range is substantially the same as the previous discourse range, but the first morpheme information belonging to the current discourse range is not interesting for the conversation control device 1. It means that there is.
[0280]
In addition, the emotional state determination unit 381 ranks the emotional degree held for the user in a higher rank when the morpheme constituting the discourse range is added to the first morpheme information by the abbreviated sentence complementing unit 350. It is also a means of attachment.
[0281]
Specifically, the abbreviated sentence complementing unit 350 selects a morpheme from the attributes (subjects consisting of principals, target cases consisting of objectives, etc.) constituting the first morpheme information extracted by the sentence structure analysis unit 430. Search for attributes that are not included. Based on the searched attribute, the abbreviated sentence complementing unit 350 adds a morpheme constituting the discourse range searched by the discourse range determining unit 340 to the attribute. When the abbreviation complementing unit 350 adds the morpheme constituting the discourse range to the first morpheme information, the emotional state determination unit 381 ranks the emotional level held by the user to a higher rank.
[0282]
That is, the input information consisting of the first morpheme information in the abbreviated sentence complementing unit 350 is the input sentence consisting of the first morpheme information before the abbreviated sentence (when the discourse range is added to the first morpheme information) Means that the information is an abbreviated sentence), the user inputs input information belonging to the current discourse range on the assumption that it belongs to the current discourse range. Therefore, the input information input at this stage tends to be an abbreviated sentence in which the morpheme consisting of the current discourse range is omitted (for details, refer to the “abbreviated sentence in the first embodiment”). (See the supplement section 350 ").
[0283]
Therefore, when the abbreviated sentence complementing unit 350 determines that the first morpheme information is an abbreviated sentence, the emotional state determination unit 381 determines that the current discourse range is substantially the same as the previous discourse range. It is determined that this first morpheme information is a cohesive factor having relevance with the previous discourse range. In this case, if the cohesion factor is a cohesion factor (+), the emotional state determination unit 381 sets the emotion level held to the user to a higher level, and if the cohesion factor is a cohesion factor (−), It is possible to set the feeling of feeling toward the user to be lower.
[0284]
Further, the emotional state determination unit 381 is also a ranking unit that ranks the emotion level held for the user based on the search result by the reflexive determination unit 320 or the reversion determination unit 330. The reflexive determination unit 320 is a fixed form obtaining unit that collates the extracted first morpheme information with each fixed form content, and searches for reflex element information including the first morpheme information from each fixed form content.
Specifically, when the reflexive determination unit 320 searches for fixed content that matches the first morpheme information from the fixed content, the emotional state determination unit 381 ranks the emotion level held by the user at a low rank. Rank.
[0285]
That is, the fact that the reflexive determination unit 320 has searched for the fixed content that matches the first morpheme information concludes that the first morpheme information is an element (interrupt factor) for interrupting the current topic. Therefore, the emotion state determination unit 381 determines that the first morpheme information is an emotion negative factor (interruption factor), and sets the emotion level held for the user to the lower one (FIG. 26, FIG. 26). 27). For example, when a user inputs “good morning” (standard content) that has nothing to do with the topic during the conversation, “good morning” becomes an interruption factor.
[0286]
Further, the return determination unit 330 is configured so that the current first morpheme information extracted by the morpheme extraction unit 410 is the past answer content extracted by the morpheme extraction unit 410 (the previous first morpheme information is the return element database 802). It is determined whether it is included in
[0287]
When the return determination unit 330 determines that the current first morpheme information is included in the past response content, the emotional state determination unit 381 returns the response to the past response content (the user answers It is possible to conclude that the first morpheme information is an interruption factor for interrupting the current topic.
[0288]
In this case, since the current first morpheme information is an emotion minus factor (interruption factor), the emotion state determination unit 381 sets the emotion level held for the user to the lower one. (See FIGS. 26 and 27).
[0289]
Further, the return determination unit 330 includes the current morpheme information extracted by the morpheme extraction unit 410 in the past first morpheme information (this past first morpheme information is stored in the return element database 802). The emotion state determination unit 381 includes the current first morpheme information in the past first morpheme information, so that the user is the same as the past input information regardless of the current topic. It can be determined that the input information is repeatedly input. In this case, the emotional state determination unit 381 determines that the input information repeatedly input is an interruption factor, and sets the emotion level held for the user to the lower one (details are given in the first embodiment). (See “Turning process” in
[0290]
When the reversal determination unit 330 or the reflexive determination unit 320 cannot retrieve the fixed content that matches the first morpheme information from the fixed content, the emotional state determination unit 381 determines the emotion level held by the user. May be ranked higher.
[0291]
Furthermore, when the input information specified by the utterance type determination unit 440 is a repulsive element such as a repulsive sentence, the emotional state determination unit 381 ranks the emotion level held by the user to a lower rank. But there is.
[0292]
Here, the repulsion element means an element for the user to repel the response content output from the conversation control device 1, and in this embodiment, for example, the input information of the user is a response. It consists of a repulsive sentence indicating that it is a repulsive sentence with respect to the contents, a negative sentence indicating that the user input information is a negative sentence with respect to the response contents, and the like.
[0293]
That is, the fact that the input type of the input information input from the user is specified by the utterance type determination unit 440 as being composed of repulsive elements means that the user and the conversation control device 1 oppose each other. Therefore, the emotion state determination unit 381 determines that the first morpheme information is an emotion negative factor (conflict factor), and sets the emotion level held to the user to a lower level ( (See FIGS. 26 and 27).
[0294]
When the topic search unit 360 does not search the second morpheme information including the first morpheme information from the second morpheme information, the understanding state determination unit 382 determines that the input information input by the user is harassed. It is a ranking means that judges that the cause is (difficult factor) and ranks the understanding level of the input information to a lower rank.
[0295]
That is, the fact that the second morpheme information including the first morpheme information is not searched from the second morpheme information in the topic search unit 360 concludes that the first morpheme information is more difficult than the second morpheme information. Therefore, the understanding state determination unit 382 determines that the first morpheme information is a difficult factor, and sets the degree of understanding of the input information to a lower level.
[0296]
The understanding state determination unit 382 can easily input information input by the user when the topic search unit 360 searches for second morpheme information including the first morpheme information from each second morpheme information. It may be determined that the degree of understanding of the input information is ranked higher. In addition, when the first morpheme information is a factor that changes the topic to the binding factor, the understanding state determination unit 382 may rank the understanding level of the input information in a higher rank.
[0297]
Here, the degree of understanding means the degree of understanding of the input information input from the user. As shown in FIG. 28, the degree of understanding can be ranked in four levels in this embodiment.
[0298]
The larger the numerical value (the direction from r (−2) to r (1)), the more the input information is understood (the direction in which the understanding is clearly extended). On the other hand, the smaller the numerical value (from r (1) to r (-2)), the more the input information is understood (the direction in which the understanding is unclear).
[0299]
Specifically, as shown in the figure, when the understanding state determination unit 382 determines that the first morpheme information is a difficult factor based on r (0), the understanding state is shifted to a lower level ( If the first morpheme information is determined to be a factor that changes the topic to a cohesive factor, the degree of understanding is shifted to a higher direction.
[0300]
When the understanding state determination unit 382 determines that the first morpheme information is an interruption factor or a conflict factor, the first morpheme information does not contribute to the understanding of the input information. Maintain the current status (see the figure). Furthermore, when the emotional level transitions in a positive direction, the understanding state determination unit 382 shifts the level of understanding level to a higher rank.
[0301]
In addition, since a difficult factor, a cohesion factor, a topic change, a discontinuation factor, and a conflict factor are the same as the content demonstrated in the term of the said emotion state determination part 381, this description is abbreviate | omitted.
[0302]
When the input information specified by the utterance type determination unit 440 is a repulsion element (oppose; also referred to as a conflict factor), the conflict state determination unit 383 ranks the user with a high degree of conflict. Is ranked. In addition, the conflict state determination unit 383 ranks the degree of conflict with the user to a lower rank when the input information specified by the utterance type determination unit 440 is not a repulsion element such as a repulsion sentence. .
[0303]
Here, the degree of confrontation means the degree of confrontation that the conversation control device 1 has against the user. As shown in FIG. 29, this degree of confrontation can be ranked in five levels in this embodiment. It means that the higher the numerical value, the lower (weaker) the degree of confrontation with the user. On the other hand, the smaller the numerical value, the higher (stronger) the degree of confrontation with the user.
[0304]
Specifically, as shown in the figure, when the conflict state determination unit 383 determines that the first morpheme information is a conflict factor with O (0) as a reference, the conflict level transitions to a lower level ( (With rank). Also, when the conflict state determination unit 383 determines that the first morpheme information is not a conflict factor but a factor that changes the discourse to a cohesion factor (not a repulsion factor), the conflict state determination unit 383 shifts the conflict level to a higher one. .
[0305]
Furthermore, when the emotion level for the first morpheme information is set in a higher direction by the emotion state determination unit 381, the conflict state determination unit 383 may set the conflict level to a lower level (a weak level of conflict). it can. In addition, when the emotion level for the first morpheme information is set in a lower direction by the emotion status determination unit 381, the conflict state determination unit 383 may set the conflict level to a higher level (a higher level of conflict). it can.
[0306]
In addition, the conflict state determination unit 383 determines that the rank of the confrontation level is as it is because the first morpheme information does not repel the response content when the first morpheme information is an interruption factor or a difficult factor. (See the figure).
[0307]
The emotion state determination unit 381, the understanding state determination unit 382, or the conflict state determination unit 383 outputs the determined emotion level, understanding level, and conflict level to the output unit 600, and the output unit receives the input emotion level and understanding level. Display the degree and degree of confrontation.
[0308]
(Ranking method using a ranking system)
The ranking method by the ranking system having the above configuration can be implemented by the following procedure. FIG. 30 is a flowchart showing the procedure of the ranking method according to this embodiment.
[0309]
As shown in the figure, first, the input unit 100 performs a step of acquiring input information input from a user (S201). And the morpheme extraction part 410 performs the step which extracts each morpheme which comprises input information based on the input information acquired by the input part 100 (S202).
[0310]
Thereafter, the phrase analysis unit 420 performs a step of grouping each morpheme into a phrase format based on each morpheme extracted by the morpheme extraction unit 410 (S203). In the present embodiment, each morpheme belonging to one phrase format is assumed to be first morpheme information.
[0311]
Next, the reflexive determination unit 320 performs a step of performing reflexive processing based on the first morpheme information specified by the phrase analysis unit 420 (S204, S205). Then, the reversal determination unit 330 performs a reversal process or a conditional reversal process based on the first morpheme information specified by the sentence structure analysis unit 430 (S206).
[0312]
After that, the discourse range determination unit 340 compares the first morpheme information with each discourse range, and performs a step of searching for a discourse range including the first morpheme information from each discourse range (S207). Furthermore, when the input information including the first morpheme information is an abbreviated sentence, the abbreviated sentence complementing unit 350 performs a step of adding a morpheme including the discourse range to the first morpheme information (S208).
[0313]
Then, the topic search unit 360 collates the first morpheme information with each second morpheme information based on the identified first morpheme information or the first morpheme information to which the morpheme is added. A step of searching for second morpheme information including the first morpheme information is performed (S209). Since the specific processes of S201 to S208 are the same as the contents of the “conversation control method” described in the first embodiment, the details of these processes are omitted.
[0314]
Next, the emotional state determination unit 381 has an emotional degree to use based on the search results obtained by the reflexive determination unit 320, the reversal determination unit 330, the discourse range determination unit 340, the abbreviated sentence complement unit 350, or the topic search unit 360. The step of ranking is performed (S210). Specifically, when the second morpheme information including the first morpheme information is not searched by the topic search unit 360, the emotion state determination unit 381 ranks the emotion level held by the user to a lower rank.
[0315]
That is, the fact that the second morpheme information including the first morpheme information is not searched by the topic search unit 360 means that each of the second morpheme information in which the first morpheme information constituting the input information input from the user is stored in advance. Therefore, the emotion state determination unit 381 determines that the input information (first morpheme information) input from the user is an emotion negative factor (difficult factor). , Set the emotion level for the user to be lower.
[0316]
On the other hand, when the second morpheme information including the first morpheme information is searched by the topic search unit 360, the emotion state determination unit 381 ranks the emotion level held for the user in a high rank.
[0317]
That is, the fact that the second morpheme information including the first morpheme information is searched by the topic search unit 360 means that each second morpheme in which the first morpheme information constituting the input information input from the user is stored in advance is stored. Since it can be determined that the information is easier than the information, the emotion state determination unit 381 determines that the input information input from the user can be understood, and has a higher feeling of feeling for the user. Set to.
[0318]
In addition, the emotional state determination unit 381 can also rank the emotion level held for the user based on the determination result by the conversation range determination unit 340. In the present embodiment, the discourse range determination unit 340 collates the first discourse range indicating the searched discourse range with the second discourse range indicating the searched discourse range, and the second discourse range is the first discourse range. It is determined whether or not there is an association with the range.
[0319]
Specifically, the emotion state determination unit 381 has a low feeling of feeling for the user when the conversation range determination unit 340 determines that the second conversation range is not related to the first conversation range. Rank to rank.
[0320]
That is, the fact that the second topic range is determined not to be related to the first topic range by the conversation range determination unit 340 means that the current conversation range (second conversation range) including the first morpheme information is the previous one. Can be determined to be different from the discourse range (the first discourse range) of (the contextual topic has no unity surname; this is also referred to as “topic change”). If the first morpheme information belonging to the second discourse range is included in the topic title (second morpheme information) associated with “low emotion level”, the first morpheme information is an emotion negative factor (topic change (−)) And the emotion level held for the user is set to be lower (see FIGS. 26 and 27).
[0321]
Thus, the topic change (-) indicates that the current conversation range is changed to another conversation range, and the first morpheme information belonging to the changed conversation range is content that is not interesting for the conversation control device 1. means.
[0322]
In this case, if the first morpheme information belonging to the second discourse range is included in the topic title (second morpheme information) associated with “high emotion level”, the emotion state determination unit 381 determines that the first morpheme information is It can be determined that it is an emotional plus factor (topic change (+)), and the emotional level held for the user can be set higher.
[0323]
As a result, topic change (+) means that the current discourse range is changed to another discourse range, and the first morpheme information belonging to the changed discourse range has interesting contents for the conversation control device 1. To do.
[0324]
On the other hand, the emotional state determination unit 381 ranks the emotion level held by the user in a high rank when the conversation range determination unit 340 determines that the second conversation range is related to the first conversation range. Attached.
[0325]
In other words, the fact that the second topic range is determined to be related to the first topic range by the conversation range determination unit 340 means that the current conversation range (second conversation range) including the first morpheme information is the previous one. Since it can be determined that it is substantially the same as the discourse range (first discourse range) (contextual topics have cohesiveness), the emotional state determination unit 381 has the second discourse range. When one morpheme information is included in the topic title (second morpheme information) associated with “high emotion level”, it is determined that the first morpheme information is an emotion plus factor (cohesion factor (+)) and used The emotion level held for the person is set higher (see FIGS. 26 and 27).
[0326]
Thus, the cohesion factor (+) is that the current conversation range is substantially the same as the previous conversation range, and the first morpheme information belonging to the current conversation range is interesting for the conversation control device 1. Means.
[0327]
In this case, when the first morpheme information belonging to the second discourse range is included in the topic title (second morpheme information) associated with “low emotion level”, the emotion state determination unit 381 determines that the first morpheme information is It can be determined that it is an emotion negative factor (cohesion factor (-)), and the emotion level held for the user can be set to be lower.
[0328]
As a result, the cohesion factor (−) is such that the current discourse range is substantially the same as the previous discourse range, but the first morpheme information belonging to the current discourse range is not interesting for the conversation control device 1. It means that there is.
[0329]
In addition, when the abbreviation complementing unit 350 adds the morpheme constituting the discourse range to the first morpheme information, the emotional state determination unit 381 ranks the emotion level held for the user to a higher rank. Can also be done.
[0330]
Specifically, the abbreviated sentence complementing unit 350 selects a morpheme from the attributes (subjects consisting of principals, target cases consisting of objectives, etc.) constituting the first morpheme information extracted by the sentence structure analysis unit 430. Search for attributes that are not included. Based on the searched attribute, the abbreviated sentence complementing unit 350 adds a morpheme constituting the discourse range searched by the discourse range determining unit 340 to the attribute. When the abbreviation complementing unit 350 adds the morpheme constituting the discourse range to the first morpheme information, the emotional state determination unit 381 ranks the emotional level held by the user to a higher rank.
[0331]
That is, the input information consisting of the first morpheme information in the abbreviated sentence complementing unit 350 is the input sentence consisting of the first morpheme information before the abbreviated sentence (when the discourse range is added to the first morpheme information) Means that the information is an abbreviated sentence), the user inputs input information belonging to the current discourse range on the assumption that it belongs to the current discourse range. Therefore, the input information input at this stage tends to be an abbreviated sentence in which the morpheme consisting of the current discourse range is omitted (for details, refer to the “abbreviated sentence in the first embodiment”). (See the supplement section 350 ").
[0332]
Therefore, when the abbreviated sentence complementing unit 350 determines that the first morpheme information is an abbreviated sentence, the emotional state determination unit 381 determines that the current discourse range is substantially the same as the previous discourse range. It is determined that this first morpheme information is a cohesive factor having relevance with the previous discourse range. In this case, if the cohesion factor is a cohesion factor (+), the emotional state determination unit 381 sets the emotion level held to the user to a higher level, and if the cohesion factor is a cohesion factor (−), It is possible to set the feeling of feeling toward the user to be lower.
[0333]
Furthermore, the emotional state determination unit 381 can also rank the emotional level held for the user based on the search result by the reflexive determination unit 320 or the reversion determination unit 330. The reflexive determination unit 320 is a fixed form obtaining unit that collates the extracted first morpheme information with each fixed form content, and searches for reflex element information including the first morpheme information from each fixed form content.
Specifically, when the reflexive determination unit 320 searches for fixed content that matches the first morpheme information from the fixed content, the emotional state determination unit 381 ranks the emotion level held by the user at a low rank. Rank.
[0334]
That is, the fact that the reflexive determination unit 320 has searched for the fixed content that matches the first morpheme information concludes that the first morpheme information is an element (interrupt factor) for interrupting the current topic. Therefore, the emotion state determination unit 381 determines that the first morpheme information is an emotion negative factor (interruption factor), and sets the emotion level held for the user to the lower one (FIG. 26, FIG. 26). 27). For example, when a user inputs “good morning” (standard content) that has nothing to do with the topic during the conversation, “good morning” becomes an interruption factor.
[0335]
Further, the return determination unit 330 is configured so that the current first morpheme information extracted by the morpheme extraction unit 410 is the past answer content extracted by the morpheme extraction unit 410 (the previous first morpheme information is the return element database 802). It is also determined whether it is included in
[0336]
When the reversal determination unit 330 determines that the current first morpheme information is included in the past response content, the emotional state determination unit 381 includes the current first morpheme information in the past response content. It is determined that the user has reverted to the previous response content (the user has re-listed to the response content), and the first morpheme information is an interruption factor for interrupting the current topic. be able to.
[0337]
In this case, since the current first morpheme information is an emotion minus factor (interruption factor), the emotion state determination unit 381 sets the emotion level held for the user to the lower one. (See FIGS. 26 and 27).
[0338]
Further, the return determination unit 330 collates the current morpheme information extracted by the morpheme extraction unit 410 with the past first morpheme information, and determines that the current first morpheme information is included in the past first morpheme information. In this case, since the current first morpheme information is included in the past first morpheme information, the emotion state determination unit 381 has the same input information as the past input information regardless of the current topic. It can be determined that the input is simply repeated. In this case, the emotional state determination unit 381 sets the emotion level held for the user to the lower one because the input information repeatedly input is an interruption factor.
[0339]
In addition, when the reflexive determination unit 320 cannot retrieve the fixed content that matches the first morpheme information from the fixed content, the emotional state determination unit 381 ranks the emotion level held by the user to a higher rank. You may attach. If the return determination unit 330 cannot retrieve past first morpheme information or past answer contents that match the current first morpheme information, the emotional state determination unit 381 ranks the emotion level held by the user with a high rank. You may rank.
[0340]
Furthermore, when the input information specified by the utterance type determination unit 440 is a repulsive element such as a repulsive sentence, the emotional state determination unit 381 ranks the emotion level held by the user to a lower rank. Can also be done.
[0341]
That is, the fact that the input type of the input information input from the user is specified by the utterance type determination unit 440 as being composed of repulsive elements means that the user and the conversation control device 1 oppose each other. Therefore, the emotion state determination unit 381 determines that the first morpheme information is an emotion negative factor (conflict factor), and sets the emotion level held to the user to a lower level ( (See FIGS. 26 and 27).
[0342]
Next, when the understanding state determination unit 382 does not search the second morpheme information including the first morpheme information from the second morpheme information by the topic search unit 360, the input information input from the user is difficult ( harass) (difficult factor), a step of ranking the degree of understanding of the input information to a lower rank is performed (S211).
[0343]
Specifically, when the second morpheme information including the first morpheme information is not searched from the second morpheme information by the topic search unit 360, the first morpheme information is more difficult than the second morpheme information. Since it can be determined, the understanding state determination unit 382 determines that the first morpheme information is a difficult factor, and sets the degree of understanding of the input information to a lower level.
[0344]
The understanding state determination unit 382 can easily input information input by the user when the topic search unit 360 searches for second morpheme information including the first morpheme information from each second morpheme information. It may be determined that the degree of understanding of the input information is ranked higher. In addition, when the first morpheme information is a factor that changes the topic to the binding factor, the understanding state determination unit 382 may rank the understanding level of the input information in a higher rank.
[0345]
Next, when the confrontation state determination unit 383 determines that the input information specified by the utterance type determination unit 440 is a repulsion element (oppose; also called a confrontation factor) such as a repulsion sentence, the degree of confrontation with the user is determined. A step of ranking to a higher rank is performed (S212). In addition, when the input information specified by the utterance type determination unit 440 is not a repulsive element such as a repulsive sentence, the confrontation state determination unit 383 ranks the degree of confrontation with the user to a lower rank. .
[0346]
Thereafter, the emotion state determination unit 381, the understanding state determination unit 382, or the conflict state determination unit 383 outputs the determined emotion level, understanding level, and conflict level to the output unit 600, and the output unit receives the input emotion level. The step of displaying the degree of understanding and the degree of confrontation is performed (S213).
[0347]
(Operations and effects of the ranking system and ranking method)
According to the invention according to the present embodiment, the reflexive determination unit 320, the reversion determination unit 330, or the topic search unit 360 collates the first morpheme information with each second morpheme information, and each second morpheme information. The second morpheme information including the first morpheme information is searched for, and the emotion state determination unit 381 can rank the emotion level held for the user based on the search result. The determination unit 381 can easily rank the emotion level held for the user by simply referring to each search result.
[0348]
As a result, the developer who develops the ranked system can easily determine the emotion level based on the search history performed between the first morpheme information and each second morpheme information. Therefore, it is not necessary to develop a complicated algorithm or a neural network for calculating the emotion level of the system with a ranked system, and as a result, the development time for developing the system can be shortened.
[0349]
The discourse range determination unit 340 has a relationship between the first discourse range related to the first morpheme information searched first and the second discourse range related to the first morpheme information searched later. And the emotional state determination unit 381 ranks the emotional level held by the user based on the result determined by the conversation range determination unit 340. Therefore, the emotional state determination unit 381 includes the first morpheme information. Can rank the feelings of the user more appropriately than the ranking of the feelings of the user based on the search results obtained by matching the second morpheme information with each other. it can.
[0350]
For example, if the second discourse range has no relationship with the first discourse range, the emotional state determination unit 381 has previously input the current input information (input from the user). Since it can be judged that the content is completely different from the input information, the user can conclude that the input is inconsistent with respect to a certain topic, and the emotion that the user has The degree can be set low (bad direction).
[0351]
On the other hand, if the second discourse range has relevance with the first discourse range, the emotional state determination unit 381 has previously input the current input information (input from the user). Since it can be determined that the input information is related, the user can conclude that the input is consistent with respect to a certain topic, and the emotional level held by the user Can be set high (good direction).
[0352]
Furthermore, when the abbreviation sentence morpheme is added to each attribute constituting the first morpheme information by the abbreviated sentence complement unit 350, the emotional state determination unit 381 is the abbreviated sentence. It can be determined that the user is inputting input information on the assumption of the previously searched discourse range.
[0353]
For this reason, when the first morpheme information consisting of input information is an abbreviated sentence, the emotional state determination unit 381 can determine that the input information is input based on the previously searched discourse range. Therefore, the user can conclude that he / she inputs the consistency of a certain topic, and can set a high emotion level (good direction) for the user.
[0354]
Furthermore, the reflexive determination unit 320 collates the first morpheme information with each fixed content, searches the fixed content including the first morpheme information from each fixed content, and the emotional state determination unit 381 Since the degree of emotion held for the user is ranked based on the search result, the emotional state determination unit 381 determines, for example, that when the fixed content including the first morpheme information is searched, the user is in a certain topic. It is possible to conclude that the input information for interrupting the user has been input, so that the feeling of feelings to the user can be ranked to a lower rank, and consequently the feelings to the user The degree can be ranked more appropriately.
[0355]
On the other hand, when the reflexive determination unit 320 cannot search for fixed content including the first morpheme information, the emotional state determination unit 381 inputs input information for interrupting a certain topic by the user. Since it can be determined that the user is answering the current topic seriously, it is possible to rank the user with a high level of emotion.
[0356]
Furthermore, when the utterance type determination unit 440 specifies an input type indicating the type of input information and the specified input type is a repulsion element such as a repulsion sentence, the emotional state determination unit 381 The emotion level held by the user can be ranked at a lower rank, and the emotion level held for the user can be ranked more appropriately.
[0357]
Furthermore, when the second morpheme information including the first morpheme information is not searched by the topic search unit 360, the understanding state determination unit 382 does not previously store the second morpheme information related to the first morpheme information. Therefore, it can be determined that the input information composed of the first morpheme information is more difficult than the sentence composed of the second morpheme information.
[0358]
For this reason, since the understanding state determination unit 382 cannot understand the input information composed of the first morpheme information, it can be determined that the user has input that the meaning is unknown to a certain topic. The degree of understanding of input information can be ranked to a lower rank.
[0359]
When the understanding state determination unit 382 can search the second morpheme information related to the first morpheme information by the topic search unit 360, the understanding state determination unit 382 determines that the user Therefore, it can be determined that the user understands the meaning of the input, and the degree of understanding of the input information can be ranked higher.
[0360]
Moreover, since the developer who develops the conversation control apparatus 1 can construct hierarchically the second morpheme information and the like for searching for the answer contents with respect to the input information of the user using a language such as DKML. The conversation control device 1 can search the database for second morpheme information related to the first morpheme information based on the first morpheme information corresponding to the user input information through a hierarchical procedure.
[0361]
That is, the reflexive determination unit 320, the reversal determination unit 330, the discourse range determination unit 340, or the topic search unit 360 includes the first morpheme information hierarchy corresponding to the input information (for example, the second morpheme information stored in the database). On the other hand, it is possible to search for appropriate second morpheme information from each second morpheme information stored in advance based on the determined hierarchy. .
[0362]
For this reason, the reflexive determination unit 320, the reversal determination unit 330, the discourse range determination unit 340, or the topic search unit 360 includes the first morpheme information composed of user input information and the pre-stored “all” second. Without matching each morpheme information, each “specific” second morpheme information that belongs to a specific hierarchy and the first morpheme information need to be collated, so the second morpheme information that approximates the first morpheme information The search can be performed in a short time, and the results can be ranked in a short time based on the search result, such as the degree of emotion held for the user.
[0363]
Furthermore, when the reversal determination unit 330 collates the current first morpheme information with the past response content and determines that the current first morpheme information is included in the past response content, the emotional state determination unit 381 It can be determined that the user is reversing the past answer contents (that the user is listening to the answer contents again).
[0364]
In this case, the emotional state determination unit 381 may determine that the user has input input information for suspending a certain topic because the user has turned over the contents of past answers. It is possible to rank the emotion level held for the user at a low rank, and as a result, the emotion level held for the user can be ranked more appropriately.
[0365]
Finally, the reversal determination unit 330 collates the current first morpheme information with the past first morpheme information, and when the current first morpheme information is included in the past first morpheme information, the emotion state determination The unit 381 can determine that the input information simply input by the user in the past is re-input regardless of the response content from the conversation control device 1.
[0366]
In this case, the emotional state determination unit 381 has nothing to do with the response content from the conversation control device 1 and the user simply inputs the same input information as the previous input information repeatedly. The emotion level held for the user can be ranked to a lower rank, and as a result, the emotion level held for the user can be more appropriately ranked.
[0367]
[Third embodiment]
A third embodiment of the present invention will be described with reference to the drawings. FIG. 31 shows the internal structure of the conversation control system according to this embodiment. As shown in the figure, the conversation control system is substantially the same as the internal structure of the conversation control device 1 (ranked device) in the second embodiment, but has a type determination unit 441 instead of the utterance type determination unit 440. The difference is that an answer type search unit 371 is provided instead of the answer sentence search unit 370 and an execution unit 390 is provided. Since it is the same as that of the structure of 1st embodiment, the modification of 1st embodiment, and 2nd embodiment except this difference, description about structures other than the said different structure is abbreviate | omitted.
[0368]
In the first embodiment, the process in which the conversation control device 1 obtains an answer sentence corresponding to the input information based on the input information from the user has been described. However, in this embodiment, the ranking is provided in the conversation control unit 300. The unit 380 ranks the emotion level and the like held for the user based on the input information from the user, and the execution unit 390 determines the input information from the user according to the ranked emotion level and the like. Acquire more appropriate response texts.
[0369]
Conceptually, as shown in FIG. 32, first, the conversation control device 1 organizes the input information based on the first morpheme information and the singular phrase included in the input information input from the user. Do. This organization means that input information is classified into an input type, a content type, and an offense and defense type, as will be described later. In the present embodiment, these input type, content type, and offense / defense type are collectively abbreviated as “sorting information”.
[0370]
Further, the singular phrase means an element obtained by removing an independent word from a character string (content uttered by a user), and as shown in FIG. 4, for example, ~ is (m2), ~ is ( m4), ~ means (m6), etc., but ~ means that it is ~, etc. Here, an independent word means a part of speech such as an adjective, a noun, or a verb in the present embodiment.
[0371]
And the conversation control apparatus 1 is based on the collation result of 1st morpheme information and each 2nd morpheme information, the feeling level (emotion state) which it has for a user, the understanding level (understanding state) with respect to input information, utilization The degree of confrontation (conflict state) is determined. In this embodiment, these emotional state, understanding state, and conflict state are collectively abbreviated as “mental state”.
[0372]
Thereafter, the conversation control device 1 grasps the input information input from the user based on the identified classification information and the determined mental state, and acquires an appropriate answer sentence based on the grasped input information.
[0373]
That is, as shown in FIG. 33, first, the conversation control device 1 collates the first morpheme information and each second morpheme information, and the second morpheme information matches the first morpheme information. Search for dimorphic information. The conversation control device 1 that has searched for the second morpheme information that matches the first morpheme information specifies the answer type associated with the searched second morpheme information, and acquires the execution function associated with the specified answer type To do.
[0374]
Then, the conversation control device 1 that has acquired the execution function, based on the classification information (input type, content type, defense type) and mental state, from among the execution elements that constitute the acquired execution function, classification information and mental Get the execution element related to the state. The conversation control apparatus 1 that has acquired the execution element executes the execution element based on the acquired execution element, and acquires an answer sentence associated with the execution element. The conversation control device 1 acquires an optimal answer sentence corresponding to the input information based on the schematic processing described above. The specific explanation is as follows.
[0375]
The type determination unit 441 determines the type of input information input from the user. In this embodiment, the type determination unit 441 includes an input type determination unit 442 and a content type determination unit 443.
[0376]
The input type determination unit 442 determines an input type indicating the type of answer content. Examples of input types include questions, questions, affirmations, and denials. The input type determination unit 442 is basically the same as the operation of the utterance type determination unit 440 in the first embodiment.
[0377]
Specifically, the input type determination unit 442 to which the sentence pattern signal is input from the phrase analysis unit 420 performs question determination, question phrase, question phrase, question phrase, negative phrase, affirmative phrase, indifference phrase, etc., as shown in FIG. Using the stored dictionary, this dictionary is compared with each morpheme included in the sentence pattern signal to determine which input type each morpheme corresponds to.
[0378]
Here, the question phrase means a character string for asking a questionable point or an unknown point. In this embodiment, a character string such as “? The question phrase means a character string for indicating that a certain event feels suspicious for truth, and in this embodiment, a character string such as “Why is the horse beautiful?” It is done.
[0379]
The negative phrase means a character string for indicating that it is negative for a certain event. In this embodiment, a negative character string or the like is given to an answer sentence from the conversation control device 1. Can be mentioned.
[0380]
The affirmative phrase means a character string for indicating that a certain event is affirmative. In the present embodiment, a positive character string or the like is given to an answer sentence from the conversation control device 1. Can be mentioned. The ambiguous phrase means a character string for indicating that it is ambiguous with respect to a certain event, and in this embodiment, a character string such as “(that is beautiful)” is cited.
[0381]
The content type determination unit 443 is a type specifying unit that specifies type information indicating a tense constituting the character string or a type of request for a certain matter based on the character string specified by the input unit 100 or the voice recognition unit 200. . In this embodiment, the content type (type information) includes, for example, types such as an action, an impression, and a mental attitude.
[0382]
As shown in FIG. 34, the content type determination unit 443 uses a dictionary storing action determination phrases, impression determination phrases, mental attitude phrases, and the like, and uses this dictionary and each morpheme and clause (determination of determination) included in the sentence pattern signal. Unit), and the content type corresponding to each morpheme is determined.
[0383]
Here, the action determination phrase is a character string for specifying a tense in which an action has occurred. In this embodiment, from the aspect and tense (past / future) such as The thing which becomes. The impression determination phrase is a character string for specifying that the other party is describing the impression. In this embodiment, “I think”, “I think”, and so on are listed.
[0384]
The mental attitude phrase means a character string for specifying a request for a certain matter and an intention for a certain matter. In the present embodiment, the phrase “I want to do”, “It seems like”, etc. are mentioned. The input type determination unit 442 and the content type determination unit 443 output the determined input type and content type to the answer type search unit 371 as a type signal.
[0385]
Based on the second morpheme information searched by the topic search unit 360, the answer type search unit 371 and each answer type associated with the second morpheme information and the input type or content type determined by the type determination unit 441 And a second search means for searching for a response type that matches the input type or content type from among the response types.
[0386]
Specifically, the answer type search unit 371 that receives the type signal from the type determination unit 441 and the search result signal from the topic search unit 360 receives the input type or content type included in the input type signal. Each answer type associated with the topic title is acquired from the conversation database 500 based on the topic title (according to the search result; second morpheme information) included in the search result signal.
[0387]
The answer type search unit 371 that acquired each answer type associated with the topic title collates the input type or content type with each answer type, and the answer that matches the input type or content type from among the answer types. The type is searched, and the execution function associated with the searched answer type is output to the execution unit 390.
[0388]
For example, as shown in FIG. 35, the answer type search unit 371 to which the search result signal is input from the topic search unit 360 and the type signal from the type determination unit 441 has a topic title corresponding to the input search result signal. If the input type corresponding to the input type signal is a conditional question sentence (CQ), the topic title (Sato; *; likes) from the conversation database 500 ) Is obtained, and the answer type (CQ) that matches the input type (CQ) is searched from among the answer types.
[0389]
The answer type search unit 371 searching for the answer type (CQ) that matches the input type (CQ) acquires the execution function 1 associated with the searched answer type (CQ), and uses the acquired execution function 1 as an execution function signal. To the execution unit 390.
[0390]
Here, the execution function means a function for obtaining an appropriate answer sentence with respect to input information from the user. In the present embodiment, the execution function includes a plurality of execution elements, and an execution unit 390 described later determines one execution element from the execution elements constituting the execution function. Based on the above, an answer sentence associated with the execution element is acquired from the conversation database 500. In this embodiment, the execution function is associated with the answer type.
[0390]
The execution unit 390 is an answer acquisition unit that acquires the answer content associated with the answer type based on the answer type searched by the answer type search unit 371. Specifically, the execution unit 390 to which the execution function signal is input from the answer type search unit 371 selects one of the execution elements included in the execution function based on the execution function corresponding to the input execution function signal. One execution element is specified, the specified execution element is executed, a corresponding answer sentence is acquired, and the acquired answer sentence is output to the output unit 600.
[0392]
As shown in FIG. 36, each execution element belonging to the execution function is roughly divided into an answer selection type corresponding to the input information or content type and an execution element (first) associated with the answer type, a ranking unit 380. It can be divided into two execution elements (second) associated with the answer selection type corresponding to the emotional state, the understanding state, or the conflicting state specified in (1).
[0393]
As shown in the figure, the first execution element can be specified based on the answer selection type and the answer type. Specifically, the execution unit 390 first specifies an answer selection type based on the input answer type. Here, the answer selection type means the type of execution element, and in this embodiment, there are types of answers, questions, etc. corresponding to the input information.
[0394]
For example, as shown in the figure, in the execution unit 390 in which the answer type “question about the condition (CQ)” is input from the answer type search unit 371, the input type corresponding to the answer type is “question”. Therefore, the answer selection type (execution element type) is specified as the “answer” corresponding to the input type “question”. The execution unit 390 that has identified the answer selection type is “condition statement (C)” because the answer selection type corresponding to the input type (CQ) is “answer” and the answer type is “condition (C)”. The “answer” for (question Q) is specified as an execution element.
[0395]
That is, since the execution element “conditional statement (C)” (question Q) “answer” is a conditional statement consisting of a question, the answer is to a conditional statement consisting of a question. Process to get the answer sentence of.
[0396]
Similarly, the execution element “result sentence” (question) shown in the figure is an answer to the result sentence including the question because the input information is the result sentence including the question. The execution element shown in the figure is the “answer” for the “time sentence” (question) “” the “answer” for the “place sentence” (question) “” the “declaration sentence” “Answer” for “question” ”“ question ”for“ question ”“ question ”” “answer” for “deny sentence” “answer” for “affirmation sentence” “ambiguous sentence” The same processing as above is performed for the “answer” for.
[0397]
Further, as shown in the figure, the execution unit 390 to which the answer type “action” (having been inputted) is input from the answer type search unit 371 has a content type corresponding to the answer type “action” “ Because it consists of “contents related to actions” (this is the type of input information entered by the user), it is necessary to further ask the user about the details of the actions related to the input information. ". The execution unit 390 specifying the answer type (action) and the answer selection type (question) performs the “ambiguous action” shown in the figure because the answer type and the answer selection type are “action” and “question”, respectively. As specified.
[0398]
The execution element “ambiguous action” is a process of acquiring an answer sentence for inquiring about an action regarding the input information because the input information has contents related to the action. Similarly, for the execution element “ambiguous impression; content related to impression (I think it is, etc.)”, processing for obtaining an answer sentence for asking the content related to the impression regarding the input information is performed. The execution element “ambiguous attitude; contents regarding mental attitude (to want to do, etc.)” performs processing for obtaining an answer sentence for asking contents regarding mental attitude regarding the input information.
[0399]
The execution unit 390 that has identified the execution element causes the identified execution element to be executed, acquires an answer sentence associated with the execution element from the conversation database 500, and outputs the acquired answer sentence to the output unit 600. A plurality of answer sentences can be stored in advance in association with execution elements.
[0400]
As a result, the execution unit 390 can identify the execution element based on the input type, and can execute the identified execution element to obtain the corresponding answer sentence. Therefore, the input information input from the user can be obtained. An appropriate answer sentence corresponding to the input type to be configured can be acquired.
[0401]
In addition, the execution unit 390 can acquire the corresponding answer sentence by specifying the execution element based on not only the input type but also the content type and executing the specified execution element. Based on the user's psychological state expressed in the input information (action, impression, mental state), an appropriate answer sentence corresponding to the input information can be obtained. As a result, the user You can enjoy the feeling of talking with other users.
[0402]
The execution unit 390, based on the second morpheme information (topic title) searched by the topic search unit 360, each priority order associated with the second morpheme information, and the rank ranked by the ranking unit 380, And rank order acquisition means for acquiring a priority order that matches the rank from among the respective priority orders. The execution unit 390 is also an answer acquisition unit that acquires the answer contents associated with the priority order based on the acquired priority order.
[0403]
Here, the second morpheme information (topic title) is associated with a plurality of answer contents to the user, and each answer contents is associated with a priority order for being selected as answer contents for the input information. A plurality of second morpheme information is stored in the conversation database 500.
[0404]
Specifically, the execution unit 390 to which a rank signal (a signal indicating the emotion level) is input from the emotion state determination unit 381 includes a rank corresponding to the input rank signal, and each priority level associated with each execution element. Are identified, a priority that matches the rank is identified from among the priorities, and an execution element associated with the identified priority is obtained. Here, the rank can be set by the process performed by the ranking unit 380 described in the second embodiment. For the detailed processing of ranking, refer to the description of the ranking unit 380 in the second embodiment.
[0405]
That is, as shown in FIG. 36, when the rank corresponding to the rank signal “transitions in the positive direction”, the execution unit 390 has increased the feeling of feeling for the user, etc. The response selection type is “strong response willingness” (in the case of a higher rank). This “strong answering motivation” means the type of answer sentence that indicates the contents for answering the input information from the user eagerly.
[0406]
The execution unit 390 that has selected “strong answer willingness” as the answer selection type identifies an execution element associated with the selected “strong answering will” (set higher rank). Therefore, the execution element associated with “strong answer willingness” means a function for obtaining an answer sentence for ambiguously answering the input information of the user.
[0407]
Similarly, when the emotion level corresponding to the rank signal has “transitioned to the standard value”, the execution unit 390 sets the answer selection type to “normal answer willingness” (when the rank is a standard value). And
[0408]
This “ordinary willingness to answer” means the type of answer text indicating the contents for normally answering the input information from the user. The execution unit 390 that has selected “normal response willingness” as the answer selection type identifies an execution element associated with the selected “normal response willingness” (set standard value rank). Therefore, the execution element associated with “normal answer willingness” means a function for obtaining an answer sentence for normally answering user input information.
[0409]
Further, the execution unit 390 sets the answer selection type to “weak answer willingness” (when the rank is a low rank) when the emotion level corresponding to the rank signal “changes to the minus direction”. . This “weak answering motivation” means the type of answer sentence indicating the contents for responding passively (or repulsively) to the input information from the user. The execution unit 390 that has selected “weak answer willingness” as the answer selection type identifies an execution element associated with the selected “weak answering willingness” (set lower rank).
[0410]
Therefore, the execution element associated with “weak answering motivation” means a function for obtaining an answer sentence for responding passively (or repulsively) to user input information. As described above, the execution unit 390 specifying the execution element. The specified execution element is executed, and an answer sentence composed of each answer selection type is acquired.
[0411]
For example, as shown in FIG. 37, when the rank size is −3 to 3 and the priority order corresponding to the size is −3 to 3, the rank signal corresponding to rank 3 is input. The execution unit 390 specifies priority 3 that matches the magnitude of rank 3 corresponding to the input rank signal, and acquires the exit element 1 associated with priority 3.
[0412]
The exit element 1 is any one of the exit elements shown in FIG. 36, and is associated with the priority in advance. The selection of the exit element is performed based on the input type or content type (see the above description). The execution unit 390 that has acquired the exit element 1 associated with the priority 3 obtains the response sentence 1-1 (the horse is really dynamic and beautiful) associated with the acquired exit element 1.
[0413]
Thereby, the execution part 390 will acquire the reply content linked | related according to the magnitude | size of this emotion degree, if the rank ranked by the ranking part 380 is made into the emotion degree which it holds with respect to a user. In this case, the corresponding response content can be acquired according to the degree of emotion held for the user.
[0414]
In addition, if the answer contents that are of interest to the user are provided according to the priority, the execution unit 390 is particularly useful for the user when the rank ranked by the ranking unit 380 is large. You can get the answers you are interested in.
[0415]
As a result, if the user does not input input information including first morpheme information that can be searched by the topic search unit 360, the topic search unit 360 searches for second morpheme information that matches the first morpheme information. Since the ranking unit 380 sets the rank to a lower rank, information that is interesting to the user cannot be acquired.
[0416]
For this reason, in order to acquire information that is of interest to the user, the user must input input information that does not set the rank ranked by the ranking unit 380 to be low. It is possible to acquire specific information while tasting the feeling of having a conversation with a person.
[0417]
The execution unit 390 is also an answer acquisition unit that does not acquire the answer contents when the rank ranked by the ranking unit 380 is the lowest. Specifically, when the rank signal is input from the ranking unit 380, the execution unit 390 is ranked in the lowest rank in emotion level, understanding level, or conflict level corresponding to the input ranked signal. Then, “Emotion control”, “Understanding control” or “Confrontation control” is specified from each execution element (see FIG. 36).
[0418]
This emotion control means that a process that does not acquire an answer sentence is performed when the emotion level of the user reaches the lowest rank. Comprehension control means that processing that does not acquire an answer sentence is performed when the degree of understanding of input information is minimized. Conflict control is a process in which an answer sentence is not acquired when the degree of confrontation with the user becomes the lowest. These emotion control, understanding control, and conflict control are collectively referred to as “mental control” (see FIG. 38).
[0419]
That is, the execution unit 390 specifying “emotional control” as an execution element does not perform a process of searching for an answer sentence from the conversation database 500 based on the specified “emotional control”. Even when the understanding control or the conflict control is specified, the execution unit 390 does not perform a process of searching for an answer sentence from the conversation database 500 as described above.
[0420]
The execution unit 390 does not acquire the response content when the response type search unit 371 searches for repulsion content including the first morpheme information when the rank ranked by the ranking unit 380 is the lowest. It is also a means.
[0421]
Here, the conversation database 500 stores a plurality of repulsion contents for repelling a certain matter in advance. In addition, the answer type search unit 371 compares the first morpheme information extracted by the morpheme extraction unit 410 with each repulsion content, and searches for the repulsion content including the first morpheme information from each repulsion content. Search means.
[0422]
For example, the input information input from the user is an emotion negative factor, the user continues to input the input information that is an emotion negative factor, and the emotion state determination unit 381 has the lowest rank regarding the emotion level. , The execution unit 390 uses the conversation database 500 to ask why the user has continued to input input information that is a negative factor of emotion based on the lowest rank. The question content (for example, “Why is this input information input?” Etc .; this is a kind of answer sentence) is acquired and the acquired question content is output (mental control; (See FIG. 38).
[0423]
Thereafter, when the execution unit 390 performs mental control, the answer type search unit 371 collates the first morpheme information input later and each repulsion content (FIG. 39), and repels the first morpheme information. Is included. Here, the content of repulsion means the content that repels the reply sentence. In this embodiment, as shown in FIG. 39, you are idiot, fool, idiot, speak, do not speak anything For example, attack phrases such as dayo. This repulsion content is stored in the conversation database 500 in advance.
[0424]
When the response type search unit 371 determines that the first morpheme information includes repulsion content, the execution unit 390 searches the conversation database 500 for a response sentence based on the determined repulsion content. Is stopped (stop process; see FIG. 38).
[0425]
Execution unit 390 performs processing that does not acquire the response content, and when the content of apology including the first morpheme information is searched by response type search unit 371, it matches the rank corresponding to the rank signal later. Based on the priority order, the response content associated with the priority order is also acquired.
[0426]
Here, the content of apology means the content for apologizing for a certain matter. In this embodiment, for example, as shown in FIG. 40, it was bad, sorry, please do not get angry. Apologies such as not saying bad things. A plurality of apology details are stored in the conversation database 500 in advance. The answer type search unit 371 compares the first morpheme information extracted by the morpheme extraction unit 410 with each apology content, and searches the apology content including the first morpheme information from each apology content. is there.
[0427]
Specifically, the execution unit 390 determines that the first morpheme information includes the apology content after stopping (stopping) the process of searching for the answer sentence from the conversation database 500. In this case, the stop process is canceled (apology / permission; see FIG. 38) based on the determined apology, and the normal state is restored (non-mental control; see FIG. 38).
[0428]
The execution unit 390 that has returned to the normal state identifies the priority that matches the rank corresponding to the rank signal based on the answer type or the like input from the answer type search unit 371 later, and is associated with the identified priority. And the acquired answer sentence is output (non-mental control; see FIG. 38).
[0429]
As a result, when the emotion level is ranked at the lowest rank, the execution unit 390 stops the process of searching for the corresponding answer sentence from the conversation database 500. As a result, the conversation control is performed. The device 1 can be used not only as a device for specifying the emotion level held by the user, but also because it does not output an answer sentence depending on the specified emotion level, etc. It can also be used as a device that lets the user plant the senses.
[0430]
In addition, when the second morpheme information input later includes the apology content, the execution unit 390 that has performed the stop process cancels the currently executed stop process. As a result, the conversation control device 1 can cancel the state in which the niece is pretending to be angry with the user, and the niece can also be used as a device whose emotion can change.
[0431]
(Conversation control method using conversation control system)
The conversation control method by the conversation control system having the above configuration can be implemented by the following procedure.
[0432]
(1) Processing to obtain an answer sentence based on the input type or content type
FIG. 41 shows a flow until the conversation control device 1 specifies the input type or content type of the input information input from the user and outputs an answer sentence based on the specified input type or content type. is there.
[0433]
As shown in the figure, first, the input unit 100 performs a step of acquiring input information input from a user (S301). And the morpheme extraction part 410 performs the step which extracts each morpheme which comprises input information based on the input information acquired by the input part 100 (S302).
[0434]
Thereafter, the phrase analysis unit 420 performs a step of grouping each morpheme into a phrase format based on each morpheme extracted by the morpheme extraction unit 410 (S303). In the present embodiment, each morpheme belonging to one phrase format is first morpheme information. Since the process so far is the same as the process (S101-S104) in 1st embodiment, detailed description of these processes is abbreviate | omitted.
[0435]
Then, the input type determination unit 442 or the content type determination unit 443 performs a step of specifying the type (input type or content type) of the input information (S304). Specifically, the input type determination unit 442 to which the sentence pattern signal is input from the phrase analysis unit 420 performs question determination, question phrase, question phrase, question phrase, negative phrase, affirmative phrase, indifference phrase, etc., as shown in FIG. Using the stored dictionary, this dictionary is compared with each morpheme included in the sentence pattern signal to determine which input type each morpheme corresponds to.
[0436]
Also, the content type determination unit 443 to which the sentence pattern signal is input from the phrase analysis unit 420 uses a dictionary storing an action determination phrase, an impression determination phrase, a mental attitude phrase, and the like, as shown in FIG. Are compared with each morpheme and clause (unit of determination) included in the sentence pattern signal to determine which content type each morpheme corresponds to. The input type determination unit 442 and the content type determination unit 443 output the determined input type and content type to the answer type search unit 371 as a type signal.
[0437]
Next, based on the second morpheme information searched by the topic search unit 360, the answer type search unit 371 and each answer type associated with the second morpheme information and the input type or content determined by the type determination unit 441 The type is collated, and an answer type that matches the input type or the content type is searched from each answer type.
[0438]
Specifically, the answer type search unit 371 that receives the type signal from the type determination unit 441 and the search result signal from the topic search unit 360 receives the input type or content type included in the input type signal. Each answer type associated with the topic title is acquired from the conversation database 500 based on the topic title (according to the search result; second morpheme information) included in the search result signal.
[0439]
Then, the answer type search unit 371 that has acquired each answer type associated with the topic title collates the input type or content type with each answer type, and matches the input type or content type from each answer type. The answer type to be searched is searched, and the execution function associated with the searched answer type is output to the execution unit 390.
[0440]
For example, as shown in FIG. 35, the answer type search unit 371 to which the search result signal is input from the topic search unit 360 and the type signal from the type determination unit 441 has a topic title corresponding to the input search result signal. If the input type corresponding to the input type signal is a conditional question sentence (CQ), the topic title (Sato; *; likes) from the conversation database 500 ) Is obtained, and the answer type (CQ) that matches the input type (CQ) is searched from among the answer types.
[0441]
Thereafter, the answer type search unit 371 searching for the answer type (CQ) that matches the input type (CQ) acquires the execution function 1 associated with the searched answer type (CQ), and executes the acquired execution function 1 It outputs to the execution part 390 as a function signal. In this embodiment, the execution function is associated with the answer type.
[0442]
Next, the execution unit 390 performs a step of specifying one execution element from the execution elements belonging to the execution function, acquiring an answer sentence associated with the specified execution element, and outputting the acquired answer sentence. (S305, S306). As shown in FIG. 36, the execution element can be specified based on the answer selection type and the answer type.
[0443]
Specifically, the execution unit 390 to which the execution function signal is input from the response type search unit 371 first specifies the response selection type based on the response type associated with the input execution function. Here, the answer selection type means the type of execution element, and in this embodiment, there are types of answers, questions, etc. corresponding to the input information.
[0444]
For example, as shown in the figure, in the execution unit 390 in which the answer type “question about the condition (CQ)” is input from the answer type search unit 371, the input type corresponding to the answer type is “question”. Therefore, the answer selection type (execution element type) is specified as the “answer” corresponding to the input type “question”. The execution unit 390 that has identified the answer selection type is “condition statement (C)” because the answer selection type corresponding to the input type (CQ) is “answer” and the answer type is “condition (C)”. The “answer” for (question Q) is specified as an execution element.
[0445]
Further, as shown in the figure, the execution unit 390 to which the answer type “action” (having been inputted) is input from the answer type search unit 371 has a content type corresponding to the answer type “action” “ Because it consists of “contents related to actions” (this is the type of input information entered by the user), it is necessary to further ask the user about the details of the actions related to the input information. ". The execution unit 390 specifying the answer type (action) and the answer selection type (question) performs the “ambiguous action” shown in the figure because the answer type and the answer selection type are “action” and “question”, respectively. As specified. Each execution element is associated with an answer selection type and an answer type. Therefore, if the answer selection type and the answer type are specified, the execution element is specified.
[0446]
Then, the execution unit 390 that has identified the execution element causes the identified execution element to be executed, obtains an answer sentence associated with the execution element from the conversation database 500, and outputs the obtained answer sentence to the output unit 600 To do.
[0447]
(2) Processing for obtaining an answer sentence based on the rank size set by the ranking unit 380
FIG. 42 shows a flow until the conversation control apparatus 1 outputs an answer sentence based on the rank size set by the ranking unit 380. Since the processes from S401 to S404 are the same as the processes from S301 to S304 described above, detailed description of these processes will be omitted.
[0448]
As shown in the figure, if the rank ranked by the ranking unit 380 is not the lowest, the execution unit 390 performs a step of outputting an answer sentence associated with this rank (S405, S406).
[0449]
Specifically, the execution unit 390 ranks the priorities associated with the second morpheme information and the ranking unit 380 based on the second morpheme information (topic title) searched by the topic search unit 360. The ranks of the ranks are collated, and a priority that matches the rank is obtained from each priority. And the execution part 390 acquires the reply content linked | related with the priority based on the acquired priority, and outputs the acquired reply content.
[0450]
For example, as shown in FIG. 37, when the rank size is −3 to 3 and the priority order corresponding to the size is −3 to 3, the rank signal corresponding to rank 3 is input. The execution unit 390 specifies priority 3 that matches the magnitude of rank 3 corresponding to the input rank signal, and acquires the exit element 1 associated with priority 3.
[0451]
The exit element 1 is any one of the exit elements shown in FIG. 36, and is associated with the priority in advance. The selection of the exit element is performed based on the input type or content type (see the above description). The execution unit 390 that has acquired the exit element 1 associated with the priority 3 obtains the response sentence 1-1 (the horse is really dynamic and beautiful) associated with the acquired exit element 1.
[0452]
On the other hand, when the rank ranked by the ranking unit 380 is the lowest, the execution unit 390 performs a step of performing a process of not acquiring the response content (mental control; S405, S407). Specifically, when the rank signal is input from the ranking unit 380, the execution unit 390 is ranked in the lowest rank in emotion level, understanding level, or conflict level corresponding to the input ranked signal. Then, “Emotion control”, “Understanding control”, or “Confrontation control” is specified from each execution element under a predetermined condition (see FIG. 36).
[0453]
This emotion control means that a process that does not acquire an answer sentence is performed when the emotion level of the user reaches the lowest rank. Comprehension control means that processing that does not acquire an answer sentence is performed when the degree of understanding of input information is minimized. Conflict control is a process in which an answer sentence is not acquired when the degree of confrontation with the user becomes the lowest. These emotion control, understanding control, and conflict control are collectively referred to as “mental control” (see FIG. 38).
[0454]
That is, when the execution unit 390 searches for the repulsion content including the first morpheme information in the answer type search unit 371 when the rank ranked by the ranking unit 380 is the lowest, the execution unit 390 Performs processing that does not acquire response contents (mental control).
[0455]
After that, the execution unit 390 performs processing that does not acquire the response content (mental control), and when the content of the apology including the first morpheme information is searched by the response type search unit 371, the rank signal is used later. Based on the priority order that matches the corresponding rank, the response content associated with the priority order is acquired (S405, S406).
[0456]
(Operation and effect of conversation control system and conversation control method)
According to the invention according to the present application having the above configuration, the answer type search unit 371 associates with the second morpheme information based on the second morpheme information including the first morpheme information constituting the input information input from the user. Each response type is compared with the content type specified by the content type determination unit 443 (the type of input information consisting of the tense constituting the character string or a request for a certain matter), and the content type is selected from each response type. Since the answer type matching the type is searched and the execution unit 390 can acquire the answer contents associated with the searched answer type, the execution unit 390 can select the type of input information input by the user. In particular, it is possible to obtain the corresponding response content based on the tense that constitutes the character string or the type of request for a certain matter, and as a result, the response suitable for the input information from the user It is possible to retrieve the contents.
[0457]
In addition, the ranking unit 380 ranks the second morpheme information according to the frequency of searching for the second morpheme information including the first morpheme information, and the execution unit 390 is searched. Based on the second morpheme information, each priority order associated with the second morpheme information is compared with the rank size ranked by the ranking unit 380, and the rank magnitude is determined from among the respective priority orders. Since the matching priority order is acquired, and the response content associated with the priority order can be acquired based on the acquired priority order, the execution unit 390 determines the rank size ranked by the ranking unit 380. , If it is an emotion level held for a user, it is possible to obtain an associated response content according to the magnitude of this emotion level. In this case, it corresponds to the emotion level held for the user Answer to It is possible to retrieve the contents.
[0458]
In addition, if the answer contents that are of interest to the user are provided according to the priority, the execution unit 390 is particularly useful for the user when the rank ranked by the ranking unit 380 is large. You can get the answers you are interested in.
[0459]
As a result, if the user does not input input information including first morpheme information that can be searched by the topic search unit 360, the topic search unit 360 searches for second morpheme information that matches the first morpheme information. Since the ranking unit 380 sets the rank to a lower rank, information that is interesting to the user cannot be acquired.
[0460]
For this reason, in order to acquire information that is of interest to the user, the user must input input information that does not set the rank ranked by the ranking unit 380 to be low. It is possible to acquire specific information while tasting the feeling of having a conversation with a person.
[0461]
Moreover, since the developer who develops the conversation control device 1 can specify, for example, ranking as an emotion level based on a search history performed between the first morpheme information and each second morpheme information. Therefore, it is not necessary to develop a complicated algorithm or a neural network for causing the conversation control device 1 to calculate the emotion level held by the user. As a result, the development time for developing the conversation control device 1 is reduced. It can be shortened.
[0462]
Further, whether the ranking unit 380 has a relationship between the first discourse range related to the first morpheme information searched first and the second discourse range related to the first morpheme information searched later. Therefore, the execution unit 390 ranks based on the search result obtained by collating the first morpheme information and each second morpheme information, and sets the priority to match the ranked size. The associated response content can be acquired, and the optimal response content corresponding to the input information from the user can be acquired.
[0463]
For example, if the second discourse range has no relationship with the first discourse range, the ranking unit 380 may input the current input information (input from the user) and the input previously input. Since it can be judged that the content is completely different from the information, the user can conclude that the input is inconsistent with respect to a certain topic. For example, the rank is set to a lower rank. be able to.
[0464]
In this case, since the rank is set low by the ranking unit 380, the execution unit 390 can output the answer content associated with the priority (low priority) that matches the low rank. For example, the user Can output information that does not show any particular interest (information with low priority) as the response content.
[0465]
On the other hand, if the second discourse range has a relationship with the first discourse range, the ranking unit 380 may input the current input information (input from the user) and the previous input. Since it can be judged that the information is relevant, the user can conclude that he / she inputs the consistency for a certain topic, for example, set the rank to a high rank. Can do.
[0466]
In this case, since the rank is set high by the ranking unit 380, the execution unit 390 can output the answer content associated with the priority (high priority) that matches the high rank. For example, the user Can be output as the contents of the answer.
[0467]
Further, the ranking unit 380 can determine that the input information including the first morpheme information is an abbreviated sentence when the discourse range morpheme is added to each attribute constituting the first morpheme information. It can be determined that the user is inputting input information on the premise of the conversation range to which the first morpheme information previously searched belongs.
[0468]
In this case, the ranking unit 380 can determine that the user is inputting input information on the premise of the discourse range to which the first morpheme information previously searched belongs, so that the user can be on a certain topic. On the other hand, it can be determined that the input is consistent, and for example, the rank can be set higher.
[0469]
In this case, the ranking unit 380 can rank more appropriately based on the first discourse range and the second discourse range, and the execution unit 390 is appropriately ranked by the ranking unit 380. Based on the size, appropriate answer contents can be acquired.
[0470]
The reflexive determination unit 320 collates the first morpheme information with each fixed content, searches the fixed content including the first morpheme information from each fixed content, and the ranking unit 380 returns the search result. The ranking unit 380 searches, for example, the fixed content including the first morpheme information, and the fixed content is the content for returning the answer sentence from the conversation control device 1 (return). In some cases, the user can determine that the input information for interrupting a certain topic has been input, and thus, for example, the rank can be set to a lower rank.
[0471]
For this reason, the ranking unit 380 can rank more appropriately based on the first morpheme information and the fixed content, and the execution unit 390 can more appropriately rank based on the appropriately ranked size. Optimal answer contents can be output. For example, if the rank is low, information that is not of interest to the user can be output as the answer contents.
[0472]
On the other hand, when the conversation control device cannot search the fixed content including the first morpheme information, the user does not input the input information for interrupting a certain topic, and the current topic is not input. On the other hand, since it can be determined that the answer is serious, the rank can be set to a higher rank, for example.
[0473]
For this reason, the ranking unit 380 can rank more appropriately based on the first morpheme information and the fixed content, and the execution unit 390 can more appropriately rank based on the appropriately ranked size. The best answer content can be output.
[0474]
In addition, when the type information specified by the input type determination unit 442 is repulsive, the ranking unit 380 sets the rank to a lower rank, so that, for example, the response content output by the user from the conversation control device 1 is displayed. On the other hand, when it is repelling, the rank can be set low.
[0475]
For this reason, the ranking unit 380 can rank more appropriately based on the type information, and the execution unit 390 outputs more optimal answer contents based on the appropriately ranked size. can do.
[0476]
Further, in the case where the current first morpheme information is included in the past response content, the ranking unit 380 returns the user to the past response content (the user is listening to the response content again). It can be determined that it is. In this case, the ranking unit 380 can conclude that the user has inputted input information for interrupting a certain topic because the user is reversing the past answer contents. For example, the rank can be set to a lower rank.
[0477]
For this reason, the ranking unit 380 can rank more appropriately based on the current first morpheme information and the past response content, and the execution unit 390 has the appropriately ranked size. Based on this, it is possible to output more optimal answer contents.
[0478]
In addition, when the current first morpheme information is included in the past first morpheme information, the ranking unit 380 does not accept the input information that the user simply input in the past as the answer content from the conversation control device. Regardless of re-input (repetitive input) can be determined.
[0479]
In this case, the ranking unit 380 has the user simply and repeatedly input the same input information as the previous input information, regardless of the answer content from the conversation control device 1. Therefore, the rank can be set to a lower rank, and for example, information that is not of interest to the user can be output as the answer content.
[0480]
Furthermore, since the execution unit 390 can prevent the content of the answer from being acquired when the rank is set to the lowest rank, as a result, the conversation control device 1 seems to be angry. It can be used as a device that allows the user to plant the senses.
[0481]
In addition, when the execution unit 390 performs processing that does not acquire the response content, when an apology content including the first morpheme information is searched, the execution unit 390 associates with the priority order based on the acquired priority order later. The execution unit 390 cancels the processing that does not acquire the response content if the content of the apology is input by the user while performing the processing that does not acquire the response content. be able to.
[0482]
As a result, the conversation control device 1 can be used not only as a device that makes the user plant an angry sensation, but also as a device that causes the user to plant a sensation that makes the user feel angry. Can also be used.
[0483]
Finally, the developer who develops the conversation control device 1 can hierarchically construct the second morpheme information and the like for searching the answer contents for the user input information using a language such as DKML. Therefore, the answer type search unit 371 or the like searches the database for the second morpheme information related to the first morpheme information based on the first morpheme information corresponding to the user input information through a hierarchical procedure. be able to.
[0484]
That is, the answer type search unit 371 or the like has a hierarchy of first morpheme information corresponding to the input information (for example, whether it is in a higher concept or a lower concept with respect to the second morpheme information stored in the database). Therefore, appropriate second morpheme information can be searched from each second morpheme information stored in advance based on the determined hierarchy.
[0485]
For this reason, the answer type search unit 371 or the like does not collate the first morpheme information composed of user input information with the previously stored “all” second morpheme information one by one. Since it is only necessary to collate each “specific” second morpheme information to which the first morpheme information belongs, the second morpheme information that approximates the first morpheme information can be searched in a short time. Answer contents related to the rank set on the basis of the rank can be quickly output.
[0486]
(Example of change)
The present invention is not limited to the third embodiment. For example, the developer who develops the conversation control device 1 can freely set the answer contents associated with the priority order or answer type according to the priority order or answer type. Note that the response content associated with the priority order or the response type is acquired by the execution unit 390.
[0487]
Thereby, the developer who develops the conversation control device 1 can freely set the answer contents to be acquired by the execution unit 390 in advance according to the priority order or the answer type. 390 obtains from the conversation database 500, for example, a saucy answer content, a self-centered answer content, a hilarious answer content, a gentle answer content, and a polite answer content with respect to the utterance content from the user. Can do. For example, in order for a developer to design the conversation control device 1 as a device for carefully answering a question from a user, the developer sets the response content expressed in a polite gradation according to priority or It should be associated with the answer type.
[0488]
As a result, the conversation control device 1 can output various answer contents according to the settings of the developer, and thus can be used as a device that gives the user an impression as if having a personality. In addition, the user can see such a conversation control device 1 as, for example, a friend, a chat friend, or a customer who has a business tie-up, and it seems as if the user is also having a conversation with another user. I can taste a sense.
[0489]
【The invention's effect】
As described above, according to the present invention, the conversation control device identifies each morpheme constituting the input information input from the user, analyzes the semantic content grasped from each identified morpheme, and performs analysis. It is possible to output the optimum answer contents corresponding to the input information from the user by acquiring the answer contents created in advance associated with the meaning contents, and to reduce the number of answer contents stored in advance. Answer contents can be output.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a schematic configuration of a conversation control device according to a first embodiment.
FIG. 2 is a block diagram showing an internal structure of a conversation control unit and a sentence analysis unit in the first embodiment.
FIG. 3 is a diagram showing the contents of each morpheme extracted by a morpheme extraction unit in the first embodiment.
FIG. 4 is a diagram showing the contents of each phrase extracted by a phrase analysis unit in the first embodiment.
FIG. 5 is a diagram showing the contents of “case” specified by the sentence structure analysis unit in the first embodiment.
FIG. 6 is a diagram showing an “uttered sentence type” specified by an utterance type determining unit in the first embodiment.
FIG. 7 is a diagram showing the contents of each dictionary stored in the utterance type database in the first embodiment.
FIG. 8 is a diagram showing the contents of a hierarchical structure built inside the conversation database in the first embodiment.
FIG. 9 is a diagram showing a detailed relationship of a hierarchical structure built inside the conversation database in the first embodiment.
FIG. 10 is a diagram showing the content of a “topic title” constructed within the conversation database in the first embodiment.
FIG. 11 is a diagram showing the content of “answer sentence type” associated with “topic title” built inside the conversation database in the first embodiment.
FIG. 12 is a diagram showing the contents of “topic title” and “answer sentence” belonging to “discourse range” built inside the conversation database in the first embodiment.
FIG. 13 is a diagram showing the contents of reflection element information stored in the reflection element database in the first embodiment.
FIG. 14 is a diagram showing the contents of a wrapping element and a morpheme of the wrapping element stored in the wrapping element database in the first embodiment.
FIG. 15 is a flowchart showing a procedure of a conversation control method according to the first embodiment.
FIG. 16 is a diagram showing utterance contents organized by a morpheme extraction unit in the first modification.
FIG. 17 is a diagram illustrating an internal configuration of a topic search unit in a second modified example.
FIG. 18 is a diagram illustrating a state in which the ratio calculation unit in the second modification collates each morpheme belonging to “case configuration” and each “topic title” for each “topic title”.
FIG. 19 is a diagram illustrating a state in which the ratio calculation unit in the second modified example collates each morpheme belonging to “each component” and each morpheme belonging to “topic title” for each “case”.
FIG. 20 is a diagram showing a schematic configuration of a conversation control system in a third modified example.
FIG. 21 is a diagram illustrating an internal configuration of a topic search unit in a sixth modified example.
FIG. 22 is a diagram showing how the topic search unit in the sixth modification collates first morpheme information with second morpheme information or an answer sentence.
FIG. 23 is a diagram showing a recording medium for storing a program in the present embodiment.
FIG. 24 is a block diagram showing an internal structure of a conversation control unit and a sentence analysis unit in the second embodiment.
FIG. 25 is a diagram showing the contents of a “discourse event table” stored in the discourse event database in the second embodiment.
FIG. 26 is a diagram showing a relationship between “topic change” and “cohesion factor” in the second embodiment.
FIG. 27 is a diagram showing the rank of emotion level determined by an emotion state determination unit in the second embodiment.
FIG. 28 is a diagram illustrating a rank of understanding level determined by an understanding state determination unit according to the second embodiment.
FIG. 29 is a diagram illustrating a rank of conflict levels determined by a conflict state determination unit according to the second embodiment.
FIG. 30 is a flowchart showing a procedure of a ranking method according to the second embodiment.
FIG. 31 is a block diagram showing an internal structure of a conversation control unit and a sentence analysis unit in the third embodiment.
FIG. 32 is a flowchart showing the flow of processing performed in the conversation control device in the third embodiment (part 1).
FIG. 33 is a flowchart showing the flow of processing performed in the conversation control device in the third embodiment (No. 2).
FIG. 34 is a diagram showing the contents of a dictionary used in the type determination unit in the third embodiment.
FIG. 35 is a diagram showing the topic title and the contents of an execution function acquired by the answer type search unit in the third embodiment.
FIG. 36 is a diagram showing the contents of execution elements determined by the execution unit in the third embodiment.
FIG. 37 is a diagram illustrating a relationship between a rank determined by a ranking unit and a priority determined by an execution unit in the third embodiment.
FIG. 38 is a diagram illustrating a flow of mental control performed by an execution unit according to the third embodiment.
FIG. 39 is a diagram showing the content of repulsion stored in the conversation database in the third embodiment.
FIG. 40 is a diagram showing apology details stored in the conversation database in the third embodiment.
FIG. 41 is a flowchart showing the procedure of a conversation control method according to the third embodiment (part 1).
42 is a flowchart showing the procedure of the conversation control method according to the third embodiment (No. 2). FIG.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Conversation control apparatus, 100 ... Input part, 200 ... Speech recognition part, 300 ... Conversation control part, 310 ... Management part, 320 ... Reflective determination part, 321 ... Ratio calculation part, 322 ... Selection part, 330 ... Reversal determination 340 ... Discourse range determination unit 350 ... Abbreviated sentence complementing unit 360 ... Topic search unit 361 ... Deletion unit 362 ... Discourse addition unit 370 ... Answer sentence search unit 371 ... Answer type search unit 380 ... Rank Appendix, 381 ... Emotion state determination unit, 382 ... Understanding state determination unit, 383 ... Conflict state determination unit, 390 ... Execution unit, 400 ... Sentence analysis unit, 410 ... Morphological extraction unit, 420 ... Phrase analysis unit, 430 ... Sentence Structure analysis unit, 440 ... utterance type determination unit, 441 ... type determination unit, 442 ... input type determination unit, 443 ... content type determination unit, 450 ... morpheme database, 460 ... utterance type database, 500 Conversation database 600 ... Output unit 700 ... Voice recognition dictionary storage unit 800 ... Communication unit 801 ... Reflection element database 802 ... Return element database 900 ... Communication unit 1000 ... Communication network 1100 ... Hard disk 1200 ... Floppy Disc, 1300 ... Compact disc, 1400 ... IC chip, 1500 ... Cassette tape

Claims (44)

The second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof is associated with a plurality of response contents to the user, and each answer content is input information input from the user. A priority storing order for selecting the answer content to each of the answer information, and a plurality of answer storage means for storing the second morpheme information in advance.
Character recognition means for identifying a character string indicating the input information based on the input information input from the user;
Based on the character string specified by the character recognition means, morpheme extraction means for extracting at least one morpheme consisting of a minimum unit constituting the character string as first morpheme information;
The first morpheme information extracted by the morpheme extraction means and the second morpheme information are collated, and second morpheme information including the first morpheme information is searched from the second morpheme information. A search means;
Based on the search result searched by the first search means, the second morpheme information is ranked according to the frequency with which the second morpheme information including the first morpheme information is searched. Ranking means to
Based on the second morpheme information searched by the first search unit, each priority order associated with the second morpheme information is compared with the ranks ranked by the ranking unit, and each priority Rank acquisition means for acquiring the priority that matches the rank from the ranks;
A conversation control system, comprising: answer acquisition means for acquiring answer contents associated with the priority order based on the priority order acquired by the rank acquisition means. The conversation control system according to claim 1,
A plurality of second morpheme information is associated with the discourse range indicating the morpheme related to the input information that will be input from the user or the answer content to the user, and each second morpheme The information is associated with a plurality of the answer contents to the user, and each answer contents is associated with the priority order for being selected as the answer contents for the input information, The answer storage means stores a plurality of the discourse ranges in advance,
A discourse search means for collating the first morpheme information extracted by the morpheme extraction means and the discourse ranges, and searching the discourse range including the first morpheme information from the discourse ranges;
The first discourse range indicating the previous discourse range searched by the discourse search means is compared with the second discourse range indicating the subsequent discourse range, and the second discourse range is related to the first discourse range. Relevance determining means for determining whether to have
The ranking control means ranks based on a determination result by the relevance determination means. The conversation control system according to claim 2,
The first morpheme information is composed of each morpheme associated with each attribute such as a subject case consisting of a main case, a target case consisting of a target case,
Based on the first morpheme information extracted by the morpheme extraction unit, an attribute search unit that searches for the attribute that does not include the morpheme from the attributes that constitute the first morpheme information;
Based on the attribute searched by the attribute search means, morpheme adding means for adding the morpheme constituting the discourse range searched by the discourse search means to the attribute,
The ranking control means ranks on the basis of a result of adding the morphemes constituting the discourse range by the morpheme adding means. The conversation control system according to claim 1,
Fixed form storage means for previously storing a plurality of fixed contents for answering fixed contents in response to the input information from the user;
Collates the first morpheme information extracted by the morpheme extracting unit and said respective fixed content, organic from the respective fixed contents, and a fixed acquisition means for retrieving said fixed content including said first morpheme information And
The ranking means is based on the search result searched by the fixed form acquisition means, according to the frequency of the fixed form contents including the first morpheme information being searched from among the fixed form contents. A conversation control system characterized by ranking. The conversation control system according to claim 4,
Having history storage means for storing the past answer contents acquired by the answer acquisition means;
The fixed form acquisition means collates the current first morpheme information acquired by the morpheme extraction means with the past response content, and determines whether the current first morpheme information is included in the past response content. Judgment,
The rank assigning means ranks based on a determination result by the fixed form obtaining means. The conversation control system according to claim 4,
The history storage means stores the past first morpheme information acquired by the fixed form acquisition means,
The fixed form acquisition unit collates the current first morpheme information acquired by the morpheme extraction unit with the past first morpheme information, and the current first morpheme information is converted into the past first morpheme information. Determine whether it is included,
The rank assigning means ranks based on a determination result by the fixed form obtaining means. The conversation control system according to claim 1,
The conversation control system, wherein the answer acquisition means does not acquire the answer contents when the rank ranked by the ranking means is lowest. The conversation control system according to claim 7,
In the answer storage means, a plurality of repulsion contents for repelling a certain matter are stored in advance,
The first search means collates the first morpheme information extracted by the morpheme extraction means with the repulsion contents, and searches the repulsion contents including the first morpheme information from the repulsion contents. And
The answer acquisition means, when the rank ranked by the ranking means is lowest, when the repulsion content including the first morpheme information is searched by the first search means, Conversation control system characterized by not acquiring. The conversation control system according to claim 8,
In the answer storage means, a plurality of apology details for apologizing for a certain matter are stored in advance,
The first search means collates the first morpheme information extracted by the morpheme extraction means with each apology content, and searches the apology content including the first morpheme information from each apology content. And
In the case where the answer acquisition unit does not acquire the answer content, and the apology content including the first morpheme information is searched by the first search unit, the priority acquired by the rank acquisition unit later A conversation control system characterized in that, based on the ranking, the answer content associated with the priority ranking is acquired. The conversation control system according to claim 1 ,
The answer storage means determines the topic requested by the user, the degree of emotion held by the user with respect to a certain matter, or the type of statement, affirmative sentence, question sentence, repulsive sentence, etc. according to the meaning content. A conversation control system characterized in that it is described in a language for hierarchically storing it in a database in association with it. The conversation control system according to claim 10 ,
The conversation control system, wherein the language is DKML. The second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof is associated with a plurality of response contents to the user, and each answer content is input information input from the user. Priority to be selected as the response content for each is associated with each,
A step of storing a plurality of the second morpheme information in advance by an answer storage means ;
Character recognition means, based on the input information input from the user, identifying a character string indicating the input information,
A step of extracting at least one morpheme consisting of a minimum unit constituting the character string as first morpheme information based on the character string specified by the character recognition unit by a morpheme extraction unit ;
The first search means collates the first morpheme information extracted by the morpheme extraction means with the second morpheme information, and includes the first morpheme information from the second morpheme information. Searching for information; and
Based on the search result searched by the first search means by the ranking means, the frequency of the search for the second morpheme information including the first morpheme information from the second morpheme information. Step to rank according to,
Based on the second morpheme information searched by the first search unit, the rank acquisition unit collates each priority rank associated with the second morpheme information with the ranked rank, and each priority rank Obtaining the priority that matches the rank from
And a step of acquiring an answer content associated with the priority order based on the priority order obtained by the order obtaining means . A conversation control method according to claim 12 , comprising:
A plurality of second morpheme information is associated with the discourse range indicating the morpheme related to the input information that will be input from the user or the answer content to the user, and each second morpheme the information, the associated plurality the answer content to users, the to each answer contents, the priority to be selected as the answer content for the input information are associated with each of the A step in which the answer storage means stores in advance a plurality of the conversation ranges;
A discourse search means for collating the first morpheme information extracted by the morpheme extraction means with the discourse ranges, and searching the discourse range including the first morpheme information from the discourse ranges; ,
The relevance determination means collates the first discourse range indicating the previous discourse range searched by the discourse search means with the second discourse range indicating the subsequent discourse range, and the second discourse range is the first discourse range. Determining whether it is relevant to the discourse range;
The ranking means includes a step of ranking based on a determination result by the relevance determination means . The conversation control method according to claim 13 , comprising:
The first morpheme information is composed of each morpheme associated with each attribute such as a subject case consisting of a main case, a target case consisting of a target case,
Based on the first morpheme information extracted by the morpheme extraction unit, an attribute search unit searches for the attribute that does not include the morpheme from the attributes that constitute the first morpheme information;
Based on the attribute searched by the attribute search means, the morpheme adding means adds the morpheme constituting the discourse range searched by the discourse search means to the attribute;
The ranking control means includes a step of ranking based on a result of adding the morphemes constituting the discourse range by the morpheme adding means . A conversation control method according to claim 12 , comprising:
A step of storing in advance a plurality of fixed contents for replying fixed contents to the input information from the user by a fixed storage means ;
A step in which a fixed form acquisition unit compares the first morpheme information extracted by the morpheme extraction unit with each fixed form content, and searches the fixed form content including the first morpheme information from the fixed form contents; ,
Based on the search result searched by the fixed form acquiring means, the ranking means ranks according to the frequency of the fixed form content including the first morpheme information from the fixed form contents. A conversation control method. The conversation control method according to claim 15 , comprising:
A history storage unit storing the past response content acquired by the response acquisition unit ;
The fixed form acquisition unit collates the current first morpheme information acquired by the morpheme extraction unit with the past response content, and whether the current first morpheme information is included in the past response content. A determining step;
The ranking means includes a step of ranking based on a determination result by the fixed form obtaining means . The conversation control method according to claim 15 , comprising:
Storing the past first morpheme information acquired by the fixed form acquisition unit in the history storage unit ;
The fixed form acquisition means collates the current first morpheme information acquired by the morpheme extraction means with the past first morpheme information, and the current first morpheme information is included in the past first morpheme information. Determining whether or not
The ranking means includes a step of ranking based on a determination result by the fixed form obtaining means . A conversation control method according to claim 12 , comprising:
The conversation control method characterized by including the step which performs the process which does not acquire the said reply content, when the said ranking of the said answer acquisition means is the lowest ranked by the said ranking means . The conversation control method according to claim 18 , comprising:
Storing in advance a plurality of repulsion contents for repelling matters in the answer storage means ;
The first search means collates the first morpheme information extracted by the morpheme extraction means with the repulsion contents, and searches the repulsion contents including the first morpheme information from the repulsion contents. Steps,
A step of not acquiring the answer content when the repulsion content including the first morpheme information is searched when the answer acquisition means has the lowest rank ranked by the ranking means; and ,
A conversation control method characterized by comprising: The conversation control method according to claim 19 ,
Storing a plurality of apology details in advance for apologizing for matters in the answer acquisition means ;
The first search means collates the first morpheme information extracted by the morpheme extraction means with the contents of each apology, and searches the contents of the apology from the contents of the apology including the first morpheme information. Steps,
In the case where the response acquisition means is performing a process that does not acquire the response content, when the apology content including the first morpheme information is searched by the first search means, it is acquired later by the rank acquisition means. Obtaining a response content associated with the priority order based on the prioritized priority order. A conversation control method according to claim 12 , comprising:
The second morpheme information of the answer storage means, the content of the answer or the priority is the topic that the user seeks, the degree of emotion held for the user with respect to a certain matter, or a statement, an affirmative sentence, a question sentence, A conversation control method characterized in that a type of repulsive sentence or the like is described in a language for hierarchically storing it in a database in association with the morpheme according to the semantic content. The conversation control method according to claim 21 ,
The conversation control method, wherein the language is DKML. The computer,
The second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof is associated with a plurality of response contents to the user, and each answer content is input information input from the user. A priority storing order for selecting the answer content to each of the answer information, and a plurality of answer storage means for storing the second morpheme information in advance .
Character recognition means for identifying a character string indicating the input information based on the input information input from the user ;
Based on the character string specified by the character recognition means, morpheme extraction means for extracting at least one morpheme consisting of a minimum unit constituting the character string as first morpheme information ;
Collating the said morpheme extracting unit extracted the first morpheme information and the respective second morpheme information, the out of the second morpheme information, the first to find second morpheme information including said first morpheme information Search means;
Based on the search result searched by the first search means, the second morpheme information is ranked according to the frequency with which the second morpheme information including the first morpheme information is searched. Ranking means to
Based on the second morpheme information searched by the first search unit, each priority order associated with the second morpheme information is compared with the ranks ranked by the ranking unit, and each priority Rank acquisition means for acquiring the priority that matches the rank from the ranks ;
A program for functioning as an answer acquisition unit that acquires the answer contents associated with the priority order based on the priority order acquired by the rank acquisition unit . The program according to claim 23 , wherein
A plurality of second morpheme information is associated with the discourse range indicating the morpheme related to the input information that will be input from the user or the answer content to the user, and each second morpheme the information, the associated plurality the answer content to users, the to each answer contents, the priority to be selected as the answer content for the input information are associated with each of the The answer storage means functions as means for storing a plurality of the conversation ranges in advance,
A discourse search means for collating the first morpheme information extracted by the morpheme extraction means and the discourse ranges, and searching the discourse range including the first morpheme information from the discourse ranges ;
The first discourse range indicating the previous discourse range searched by the discourse search means is compared with the second discourse range indicating the subsequent discourse range, and the second discourse range is related to the first discourse range. Function as a relevance judging means for judging whether or not
The unit with the rank, the program for causing a function as a means for ranked based on the determination result. A program according to claim 24 , wherein
The first morpheme information is composed of each morpheme associated with each attribute such as a subject case consisting of a main case, a target case consisting of a target case,
Based on the first morpheme information extracted by the morpheme extraction unit, an attribute search unit that searches for the attribute that does not include the morpheme from the attributes that constitute the first morpheme information ;
Based on the attribute searched by the attribute search means, function as morpheme adding means for adding the morpheme constituting the discourse range searched by the discourse search means to the attribute ,
Program for functioning as means for ranked on the basis of a result of the morphemes constituting the discourse scope means with the rank is added. The program according to claim 23 , wherein
Fixed form storage means for previously storing a plurality of fixed contents for answering fixed contents in response to the input information from the user ;
The first morpheme information extracted by the morpheme extraction unit is compared with each fixed content, and the fixed content including the first morpheme information is searched from the fixed content to function as a fixed acquisition unit. ,
Based on the search result searched by the fixed form acquiring means, the ranking means is selected according to the frequency of the fixed form content including the first morpheme information from the fixed form contents. A program for functioning as a means of ranking. The program according to claim 26 , wherein
Function as history storage means for storing the past answer contents acquired by the answer means ;
The fixed form acquisition unit compares the current first morpheme information acquired by the morpheme extraction unit with the past response content, and whether the current first morpheme information is included in the past response content. Function as a means to determine,
A program for causing the ranking means to function as a means for ranking based on a determination result by the fixed form obtaining means . The program according to claim 26 , wherein
Wherein the history storage means, the first morpheme information in the past acquired by the standard acquisition means is stored,
The first acquisition unit compares the current first morpheme information acquired by the morpheme extraction unit with the previous first morpheme information, and the current first morpheme information is included in the past first morpheme information. Function as a means to determine whether
A program for causing the ranking means to function as a means for ranking based on a determination result by the morpheme extraction means . The program according to claim 23 , wherein
A program for causing the answer acquisition means to function as means for not acquiring the answer contents when the rank ranked by the ranking means is lowest. A program according to claim 29 ,
In the answer storage means, a plurality of repulsion contents for repelling a certain matter are stored in advance,
The first search means collates the first morpheme information extracted by the morpheme extraction means with the repulsion contents, and searches the repulsion contents including the first morpheme information from the repulsion contents. Function as a means to
When the repulsion content including the first morpheme information is searched for by the first search means when the rank ranked by the ranking means is the lowest in the response acquisition means, A program for functioning as a means not to acquire. A program according to claim 30 , wherein
In the answer storage means, a plurality of apology details for apologizing for a certain matter are stored in advance ,
In the first search means, the first morpheme information extracted by the morpheme extraction means and the contents of each apology are collated, and the apology contents including the first morpheme information are searched from the contents of each apology. Function as a means to
In the case where the processing for not acquiring the content of the response is performed in the response acquisition unit, when the content of the apology including the first morpheme information is searched for by the first search unit, The program for functioning as a means to acquire the reply content linked | related with this priority based on the acquired said priority. The program according to claim 23 , wherein
The second morpheme information of the answer storage means, the content of the answer or the priority is the topic that the user seeks, the degree of emotion held for the user with respect to a certain matter, or a statement, an affirmative sentence, a question sentence, A program characterized in that a type of repulsive sentence is described in a language for hierarchically storing it in a database in association with the morpheme according to its semantic content. A program according to claim 32 , wherein
The program is characterized in that the language is DKML. The computer,
The second morpheme information indicating a morpheme composed of one character, a plurality of character strings, or a combination thereof is associated with a plurality of response contents to the user, and each answer content is input information input from the user. A priority storing order for selecting the answer content to each of the answer information, and a plurality of answer storage means for storing the second morpheme information in advance .
Character recognition means for identifying a character string indicating the input information based on the input information input from the user ;
Based on the character string specified by the character recognition means, morpheme extraction means for extracting at least one morpheme consisting of a minimum unit constituting the character string as first morpheme information ;
The first morpheme information extracted by the morpheme extraction means and the second morpheme information are collated, and second morpheme information including the first morpheme information is searched from the second morpheme information . A search means;
Based on the search result searched by the first search means, the second morpheme information is ranked according to the frequency with which the second morpheme information including the first morpheme information is searched. Ranking means to
Based on the second morpheme information searched by the first search unit, each priority order associated with the second morpheme information is compared with the ranks ranked by the ranking unit, and each priority Rank acquisition means for acquiring the priority that matches the rank from the ranks ;
A computer-readable recording medium storing a program for functioning as an answer acquisition unit for acquiring the answer contents associated with the priority order based on the priority order acquired by the rank acquisition unit . A recording medium according to claim 34 ,
A plurality of second morpheme information is associated with the discourse range indicating the morpheme related to the input information that will be input from the user or the answer content to the user, and each second morpheme the information, the associated plurality the answer content to users, the to each answer contents, the priority to be selected as the answer content for the input information are associated with each of the The answer storage means functions as means for storing a plurality of the conversation ranges in advance ,
A discourse search means for collating the first morpheme information extracted by the morpheme extraction means and the discourse ranges, and searching the discourse range including the first morpheme information from the discourse ranges ;
The first discourse range indicating the previous discourse range searched by the discourse search means is compared with the second discourse range indicating the subsequent discourse range, and the second discourse range is related to the first discourse range. Function as a relevance judging means for judging whether or not
Means with the rank determination result computer-readable recording medium storing a program to function as means for ranked based. The recording medium according to claim 35 , wherein
The first morpheme information is composed of each morpheme associated with each attribute such as a subject case consisting of a main case, a target case consisting of a target case,
Based on the first morpheme information extracted by the morpheme extraction unit, an attribute search unit that searches for the attribute that does not include the morpheme from the attributes that constitute the first morpheme information ;
Based on the attribute searched by the attribute search means, function as morpheme adding means for adding the morpheme constituting the searched discourse range to the attribute ,
A computer-readable recording medium storing a program for causing the ranking means to function as a means for ranking based on a result of adding the morpheme constituting the discourse range by the morpheme adding means . A recording medium according to claim 34 ,
Fixed form storage means for previously storing a plurality of fixed contents for answering fixed contents in response to the input information from the user ;
The first morpheme information extracted by the morpheme extraction unit is compared with each fixed content, and the fixed content including the first morpheme information is searched from the fixed content to function as a fixed acquisition unit. ,
According to the magnitude of the frequency at which the fixed content including the first morpheme information is searched from the fixed content based on the search result searched by the fixed form acquiring unit to the ranking unit. A computer-readable recording medium having recorded thereon a program for causing it to function as a ranking means . A recording medium according to claim 37 , wherein
Function as history storage means for storing the past response contents acquired by the response acquisition means ;
In the fixed form acquisition means, the current first morpheme information acquired by the morpheme extraction means is compared with the past response content, and whether the current first morpheme information is included in the past response content. Function as a means to determine ,
A computer-readable recording medium storing a program for causing the ranking means to function as a ranking means based on a determination result by the fixed form obtaining means . A recording medium according to claim 37 , wherein
The first morpheme information in the past obtained is stored in the history storing means,
The fixed form acquisition unit compares the current first morpheme information acquired by the morpheme extraction unit with the previous first morpheme information, and the current first morpheme information is converted into the past first morpheme information. Function as a means to determine whether it is included ,
A computer-readable recording medium storing a program for causing the ranking means to function as a ranking means based on a determination result by the fixed form obtaining means . A recording medium according to claim 34 ,
A computer-readable recording medium storing a program for causing the answer obtaining means to function as means for not obtaining the answer contents when the rank ranked by the ranking means is lowest. The recording medium according to claim 40 , wherein
In the answer storage means, a plurality of repulsion contents for repelling a certain matter are stored in advance ,
The first search means collates the first morpheme information extracted by the morpheme detection means with each repulsion content, and retrieves the repulsion content including the first morpheme information from each repulsion content Function as a means to
When the repulsion content including the first morpheme information is searched for by the first search means when the rank ranked by the ranking means is the lowest in the response acquisition means, a computer-readable recording medium storing a program order to function as an acquisition and not means. The recording medium according to claim 41 , wherein
In the answer storage means, a plurality of apology details for apologizing for a certain matter are stored in advance ,
In the first search means, the first morpheme information extracted by the morpheme extraction means and the contents of each apology are collated, and the apology contents including the first morpheme information are searched from the contents of each apology. Function as a means to
The answer storage means, when you have made no acquisition process the answer content, the when the apology content including the first morpheme information is retrieved by the first search means, after the order acquisition means The computer-readable recording medium which recorded the program for functioning as a means to acquire the reply content linked | related with this priority order based on the said priority order acquired by (4). A recording medium according to claim 34 ,
The second morpheme information of the answer acquisition means, the answer content or the priority order is the topic that the user seeks, the feeling of the user with respect to a certain matter, or a statement, an affirmative sentence, a question sentence, A computer-readable recording medium on which a program is recorded, characterized in that a type of repulsive sentence or the like is described in a language for hierarchically storing it in a database in association with the morpheme according to its semantic content. A recording medium according to claim 43 , wherein
A computer-readable recording medium storing a program, wherein the language is DKML.

Images