JP2810750B2 - Language training equipment - Google Patents

Description

Description: FIELD OF THE INVENTION The present invention relates to a language training apparatus utilizing a computer system mainly for conversation simulation,
In particular, it is intended to perform conversation control corresponding to the learner's language ability level to appropriately train the learner.

[Prior art] It has been pointed out that language education using grammar and transliteration pedagogy is insufficient in terms of conversation, and audio tapes, CDs, Video tapes and the like are commercially available. However, although they are effective for memorizing many example sentences and for memorizing patterned conversations, they have the ability to respond to situations that are the most important factors in real conversations, situational responses, intentional responses, etc. There are many cases where training is difficult.

In order to solve such inconveniences, a language training apparatus centered on a conversation simulation and using a computer system has been proposed (for example, "an intelligent CAI system for language training based on a conversation simulation"). Structure, Hideki Yamamoto et al., Transactions of Information Processing Society of Japan, Vol. 30, No. 7, pp 980-917, July 1989).

[Problems to be Solved by the Invention] However, in a conventional language training apparatus centered on a conversation simulation, uniform conversation control is performed regardless of a learner's language ability level. If the level of training using the language training device is low, it is unsatisfactory for learners with a high level of language skills, and if the level of training with the language training device is high, learners with a low level of language skills may not be able to keep up with learning. That is, there is a problem that language training with appropriate guidance is not performed.

In order to solve such inconvenience using a conventional language training apparatus, it is necessary to separately prepare language training apparatuses having different training levels in consideration of the learner level.

The present invention has been made in view of the above points,
By the conversation control corresponding to the learner's language level, it is possible to perform dialogue training with an appropriate difficulty level. It is an object of the present invention to provide a language training apparatus having high flexibility with respect to a training level.

[Means for Solving the Problem] In order to solve the problem, according to the present invention, in a language training apparatus centered on a conversation simulation for forming an output sentence in response to an input sentence from a learner, based on an input sentence Language level detecting means for detecting the language level of the learner, and output sentence determining means for determining the output sentence according to the detected language level in addition to the semantic expression of the input sentence.

[Operation] In the present invention, the language skill level detecting means detects the learner's language skill level based on the input sentence, and gives it to the output sentence determining means. The output sentence determining means determines the output sentence according to the detected language ability level in addition to the semantic expression of the input sentence.

In this way, the training takes into account the learner's language level.

[Embodiment] An embodiment of the present invention applied to English training will be described below with reference to the drawings.

Here, FIG. 1 is a flowchart for explaining the operation in this embodiment, FIG. 2 is a configuration diagram when the language training apparatus of this embodiment is viewed as a computer system, and FIG.
FIG. 4 is a functional block diagram of the apparatus of the embodiment, FIG. 4 is an explanatory view of an example of a conversation control rule of the embodiment, FIG. 5 is an explanatory view showing an example of a dialogue between the language training apparatus and a learner, and FIG. FIG. 7 is an explanatory diagram showing a part of the grammar dictionary knowledge of the example, and FIG. 7 is an explanatory diagram showing an example of a parse tree according to the embodiment.

In FIG. 2, this language teaching device using a computer system is, like other computer systems, also a central processing unit (CPU) 1, a main memory 2, a display device 3, a keyboard device 4, and a hard disk. The device 5 is provided.

The hard disk device 5 includes a natural language processing program related to input sentence understanding and output sentence generation processing, a conversation control program related to output sentence determination processing based on the input sentence, and various types of processing used in each stage. Knowledge is stored. For example, a conversation control rule shown in FIG. 4 and a word dictionary and grammar rules shown in FIG. The natural language processing program and the conversation control program are loaded into the main memory 2 and executed by the CPU 1. As is well known, the display device 3 and the keyboard device 4 are for a man-machine interface between the learner and the training device, and a dialogue is executed via these devices 3 and 4.

This embodiment is characterized in that conversation control corresponding to a learner's language ability level is performed.
Next, the conversation control process corresponding to the learner's language skill level will be described with reference to the functional configuration shown in FIG. 3 and the flowchart shown in FIG. FIG. 1 shows processing when one input sentence is given, and is repeated every time an input sentence is given.

When a certain input sentence is given, the input sentence understanding unit 6 performs a process for understanding the meaning of the input sentence using the grammar dictionary knowledge 9. That is, the sentence understanding unit 6 parses a sentence input by the learner to generate a parse tree (see FIG. 7), and extracts information that can detect the learner's language ability level (step 12). ).

At this time, the dictionary and grammar rules used are grammar dictionary knowledge 9
Is stored in In this embodiment, each word in the dictionary describes a level indicating the level of difficulty of the word, and each grammar rule describes a level indicating the level of difficulty of the rule. The input sentence comprehension unit 6 analyzes the input sentence and, at the same time, stores all the words used and the levels described in the grammatical rules in a variable in order to detect the language skill level of the learner. Further, knowledge about grammatical errors is also stored in the grammar dictionary knowledge 9, and the input sentence understanding unit 6
If an error is found during the syntax analysis, an educational message is presented, and the number of errors is substituted into a variable in order to provide for the detection of the learner's language ability level as described above.
Misspelling of words is algorithmically detected during parsing, but when a spelling is found, an educational message is presented and the number of errors is substituted into a variable as described above. Further, the number of words in the input sentence is also substituted into a variable for use in detecting the learner's language ability level.

Next, the input sentence comprehension unit 6 generates a semantic expression from the parse tree obtained by analysis and passes it to the conversation control unit 7 (step 1).
3).

In the conversation control unit 7, first, the use word level of the learner,
Using the variable to which the used grammar rule level, the number of errors, and the number of words are substituted, a comprehensive learner's language ability level is calculated based on a predetermined formula (step 14). The conversation control knowledge 10 stores conversation control rules (see FIG. 4). The conversation control rule is a rule in accordance with, for example, an if-then format that defines the next utterance of the language training device based on the semantic expression of the input sentence of the learner and the language ability level of the learner.
When the language ability level of the learner is detected by calculation as described above, the conversation control unit 7 determines the content to be uttered next by the device according to the conversation control rule (step S1).
15). The determined utterance content is provided to the output sentence generation unit 8 together with the learner's language ability level information.

The output sentence generation unit 8 receives the determined utterance content and the language ability level information, generates an output sentence based on the output sentence generation knowledge 11, and outputs and outputs the output sentence (step 16). When generating the output sentence, if the learner's language skill level is high,
A sentence using a word or syntax with a high degree of difficulty is selected.

In this way, conversation control according to the learner's language ability level is realized.

Next, a specific example of conversation control according to the embodiment will be described with reference to the drawings. FIG. 5 shows an example of a dialogue performed by applying the conversation control rule shown in FIG. 4, and with reference to FIG. 5 and with reference to FIG. 4, FIG. 6, and FIG. A specific example will be described.

The conversation control rules 21 to 28 shown in FIG. 4 are rules in accordance with the if-then format as described above, and each include an if part (condition part) 17 and a then part 18. In the case of this embodiment, unlike the conventional case, the if part 17
In addition to 19, it also contains the learner's language level information 20. Therefore, the then section 18 that satisfies the conditions of the semantic expression 19 of the input sentence and the learner's language ability level information 20, that is, the utterance sentence (output sentence) of the device is selected.

In FIG. 5, a speaker 29 indicates a sentence provider side, "S" indicates a training device, "Gaku" indicates a learner, and numbers indicate the order of speech. In this example, the training device is an example of a hotel receptionist, and a learner wants to stay. This is a dialogue with the display output of the system and the key input of the learner. When the speaker 29 is a learner, the utterance content 30 is associated with variables 31 to 33 used for detecting the learner's language ability level. The first variable * level * is the first variable
A list of the level of each word whose element (the first information in the figure in parentheses) is used in the input sentence, and each element whose second element (the second information in the figure in parenthesis) is used in the input sentence Shows a list of grammar rule levels. The second variable * error *
Indicates the number of grammatical errors, and the third variable * words * indicates the number of words in the sentence.

FIG. 6B shows the contents 51 to 61 of the grammar dictionary knowledge 9 for some of the words used in the utterance contents (input sentences) 30 of the learner in this dialogue example. (A) shows the contents 41 to 49 of the grammar dictionary knowledge 9 on the grammar rules (phrase structure rules in the vocabulary function grammar) used for the learner's input sentence in this dialogue example. These dictionary contents are used for determining the first variable * level *.

FIGS. 7 (A) to 7 (C) each show a parse tree of the utterance content (input sentence) 30 of the learner in the dialogue example of FIG. 5, and the first variable * leve
The second element of l * is determined.

First, at the start of the conversation, the training device makes the first utterance “Ma
y I help you? "(Fig. 5c1). In response, the learner responds with" I want stay heree. "(Study 2).

The input sentence understanding unit 6 understands that the word level of the input sentence is (2222) from the dictionary contents 51 to 54 of the words shown in FIG. 6 (B). That is, (2222) is substituted for the first element of the first variable * level *. The input sentence comprehension unit 6 forms a parse tree for the input sentence as shown in FIG. 7 (A), and determines the level for the phrase structure rule contained therein in FIG. 6 (A). Extract from the contents 45, 43, 49, and 48 of the grammar rules shown in ()). For example, the phrase structure rule that s (sentence) of the input sentence is composed of np (noun phrase) and vp (verb phrase) is associated with level 2 from the rule with reference numeral 45 in FIG. 6 (B). . By performing similar processing, (2222) is substituted for the second element of the first variable * level *.

In the second variable * error *, 2 is substituted because this input sentence has an error without the infinitive to and an error in spelling here. The word number 4 is substituted for the third variable * words *.

The input sentence comprehension unit 6 obtains the variables 31 to 33 thus obtained.
Is given to the conversation control unit 7. Of course, the input sentence "I want stay
"I want to stay" as an input meaning expression of "heree." is also given to the conversation control unit 7.

The conversation control unit 7 first detects the learner's language ability level from the given variables. The calculation for detection is calculated by the following equation: {(total of elements of * level *) − (* error * × 2) + (* word * −3)} {(number of elements of * level *). ). The range to which the value obtained in this equation (1) (in this case, 1.67) belongs is detected, and the language skill level of the learner is detected.

It should be noted that a learner with a higher language ability level has a higher used word level and a higher grammatical rule level, a larger number of words constituting a sentence, and fewer errors than a learner with a lower language ability. (1) Formula is defined.

It is assumed that the result of the learner level of 1 is obtained for the first input sentence "I want stay heree." The conversation control unit 7
If the semantic expression of the input sentence is "I want to stay" and language level 1
From the rule 22 in the part 17, "Single, or
"double, or twin?" is described here in a natural language, but is actually different, and the output sentence generation unit 8 to which the decision information is given from the conversation control unit 7 is uttered in the natural language (FIG. 5). Perform step 3).

In response to this, it is assumed that the learner responded with “Twin.” (FIG. 5, Study 4). In response to this, the input sentence understanding unit 6 and the conversation control unit 7 perform the same processing as described above. The first element of the first variable * level * is (2), the second element of the first variable * level * is (12), the second variable * error * is 0, and the third variable * words * is 1 and the value according to the equation (1) is 1. Thus, the learner's language ability level is determined to be the same as before, and the device is called “Do you have a reservation?” (FIG. 5c5) based on rule 25 that also uses this language ability level as one condition. Speak.

In response, the learner responded “Yes, I have a reservati
on. "(FIG. 5). In response to this, the input sentence understanding unit 6 and the conversation control unit 7 perform the same processing as described above. One element is (23222
3), the second element of the first variable * level * is (2223),
The variable * error * is 0, the third variable * words * is 5, and the value according to equation (1) is 2.5. Thus, the language proficiency level is determined to be 2 which is larger by 1 than before, and the apparatus makes the next utterance (FIG. 5, C7) based on the rule 27 with this language proficiency level 2 as one condition.

Therefore, according to the above-described embodiment, the next utterance of the language training apparatus is determined by detecting the learner's language ability level, so that training appropriate for the learner's language ability level is appropriately performed. And the learning effect can be improved.

Also, appropriate training can be performed for learners of different language proficiency levels using the same training device.

Furthermore, according to the above-described embodiment, since the language level is detected every time an input sentence is given, high training is performed in a situation where the learner is good, and low training is performed in a situation where the learner is not good. And highly flexible training can be conducted.

In the above description, the present invention is applied to a language training system for English conversation, but the present invention can of course be applied to other language systems.

Also, the present invention can be applied to a language training apparatus in which an input sentence and an output sentence are spoken.

Further, as a viewpoint for detecting the language proficiency level, not only the word level, the grammar rule level, the number of errors, and the number of words used, but also other viewpoints may be used, and all of the above four viewpoints may be used. The language proficiency level may be determined using only a part of the viewpoint without using the language. The calculation formula for determining the language proficiency level is not limited to the formula (1).

In the above embodiment, the language skill level is reviewed for each input sentence. However, the language skill level may be reviewed for a plurality of input sentences.

[Effects of the Invention] As described above, according to the present invention, the apparatus prepares utterances of different levels as utterances of the same content, and when an input sentence is given, the linguistic ability of the learner along with the semantic expression Since the level is detected and the utterance of the device is determined from the semantic expression and the language ability level, training can be performed in accordance with the learner's language ability level, and the learning effect can be improved.

[Brief description of the drawings]

FIG. 1 is a flowchart for explaining the operation of an embodiment of the language training apparatus according to the present invention. FIG. 2 is a block diagram of the language training apparatus of this embodiment when viewed as a computer system. FIG. 4 is a functional block diagram of the device, FIG. 4 is an explanatory diagram of an example of a conversation control rule of the embodiment, FIG. 5 is an explanatory diagram showing an example of a dialogue between the language training device and a learner, and FIG. 6 is a grammar dictionary of the embodiment Explanatory diagram showing a part of knowledge, 7th
The figure is an explanatory diagram showing an example of a parse tree according to the embodiment. 6 input sentence understanding unit 7 conversation control unit 8 output generation unit 9 grammar dictionary knowledge 10 conversation control knowledge 11
... output sentence generation knowledge, 19 ... input semantic expression, 20 ... language skill level.

──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Yukari Sawayama 3-8-10 Uchiyama, Chikusa-ku, Nagoya-shi, Aichi Prefecture Inside the Oki Techno Systems Laboratories (72) Inventor Satoko Kai 1-7-7 Toranomon, Minato-ku, Tokyo No.12 Oki Electric Industry Co., Ltd. (72) Inventor Mariko Osato 1-7-12 Toranomon, Minato-ku, Tokyo (72) Inventor Hideki Yamamoto 1-7-7 Toranomon, Minato-ku, Tokyo No. 12 Oki Electric Industry Co., Ltd. (58) Field surveyed (Int. Cl. ⁶ , DB name) G09B 19/06 G09B 5/00

Claims (1)

(57) [Claims] 1. A language training apparatus centered on a conversation simulation for forming an output sentence according to an input sentence from a learner, wherein the language skill level detecting means detects a learner's language ability level based on the input sentence. And an output sentence determining means for determining an output sentence according to the detected language ability level in addition to the semantic expression of the input sentence.