CN109727591B - Voice search method and device - Google Patents
Voice search method and device Download PDFInfo
- Publication number
- CN109727591B CN109727591B CN201811645648.9A CN201811645648A CN109727591B CN 109727591 B CN109727591 B CN 109727591B CN 201811645648 A CN201811645648 A CN 201811645648A CN 109727591 B CN109727591 B CN 109727591B
- Authority
- CN
- China
- Prior art keywords
- voice
- search
- module
- resources
- resource
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Machine Translation (AREA)
Abstract
The invention provides a method and a device for voice search, wherein the method comprises the following steps: establishing an inverted index of the resource in advance, and updating the heat of the resource; collecting user voice data; recognizing the user voice data and converting the user voice data into text data; performing semantic analysis on the text data to obtain core keywords; obtaining a preliminary search result by adopting different voice search strategies according to the length of the core keyword; and sequencing the initial search results according to the popularity of the resources to obtain a final voice search result. The invention solves the problems of frequent error, slow search speed and poor user voice search experience of the existing voice recognition, provides a voice search technical scheme which can quickly and accurately output search results within a certain error recognition range, and can greatly improve the user voice search experience effect.
Description
Technical Field
The invention belongs to the technical field of voice search, and particularly relates to a voice search method and device.
Background
With the development of intelligent devices, human-computer interaction, and Al technologies, requirements for voice search related technologies, particularly in terms of speed and accuracy, are becoming more and more stringent. At present, voice search generally comprises four steps of voice recognition, semantic analysis, resource search and result sorting. Due to factors such as polyphones, accents and dialects of users, environmental noise and the like, voice recognition often makes mistakes, so that the search space expands, the search result and the user intention are greatly different, the search speed is slow, and the voice search experience of the users is greatly influenced. The voice search technology provided by the text can quickly and accurately output the search result within a certain error recognition range, and can greatly improve the voice search experience effect of a user.
Disclosure of Invention
In view of this, embodiments of the present invention provide a method and an apparatus for voice search, so as to solve the problems of frequent error in voice recognition, slow search speed, and poor user voice search experience in the prior art, and achieve fast and accurate output of search results within a certain range of false recognition, and greatly improve the user voice search experience effect.
In a first aspect, a method for voice search is provided, where the method includes:
establishing an inverted index of the resource in advance, and updating the heat of the resource;
collecting user voice data;
recognizing the user voice data and converting the user voice data into text data;
performing semantic analysis on the text data to obtain core keywords;
obtaining a preliminary search result by adopting different voice search strategies according to the length of the core keyword; and sequencing the initial search results according to the popularity of the resources to obtain a final voice search result.
Further, the pre-establishing the reverse index of the resource specifically includes:
numbering all resources to be searched and generating a unique ID of a resource index;
extracting key words and sentences used for retrieving the resources according to the content of the resources;
performing word segmentation, and segmenting the keywords and the sentences into single Chinese characters to obtain a keyword set;
and establishing an inverted index according to a corresponding strategy according to the size of the keyword set.
Further, the establishing an inverted index according to the size of the keyword set and the corresponding policy specifically includes:
judging the value of the size K of the keyword set;
when K is less than or equal to 2, acquiring the pronunciation of each word in the keyword set to obtain a pronunciation set { a1, a 2.. an }, and establishing a corresponding pinyin resource index for each pronunciation in the pronunciation set;
and when K is larger than 2, acquiring the code of each word in the keyword set to obtain a character code set { c1, c 2.. once, cn }, and establishing a corresponding character resource index for each character in the character code set.
Further, obtaining a preliminary search result by adopting different voice search strategies according to the length of the core keyword; and ranking the initial search results according to the popularity of the resources to obtain a final voice search result, which specifically comprises the following steps:
segmenting the obtained core keywords to obtain a user search keyword set;
judging the value of the size M of the user search keyword set;
when M is less than or equal to 2, searching the pinyin resource index to obtain a primary search result; sequencing the initial search results according to the popularity of the resources to obtain a final voice search result;
when M is larger than 2, searching the character resource index, and adopting different fault-tolerant strategies according to the value length of M to obtain a primary search result; and sequencing the initial search results according to the popularity of the resources to obtain a final voice search result.
In a second aspect, an apparatus for voice search is provided, the apparatus comprising:
the preprocessing module is used for pre-establishing an inverted index of the resource and updating the heat of the resource;
the voice data acquisition module is used for acquiring user voice data;
the voice recognition module is used for recognizing the user voice data and converting the user voice data into text data;
the semantic analysis module is used for performing semantic analysis on the text data to obtain core keywords;
the voice search and result output module is used for obtaining a preliminary search result by adopting different voice search strategies according to the length of the core keyword; and sequencing the initial search results according to the popularity of the resources to obtain a final voice search result.
Further, the preprocessing module device further comprises:
the index ID generation module is used for numbering all resources to be searched and generating the unique ID of the resource index;
the keyword sentence extraction module is used for extracting keywords and sentences used for retrieving the resources according to the content of the resources;
the word segmentation module is used for segmenting words and the keywords and the sentences into single Chinese characters to obtain a keyword set;
and the reverse index establishing module is used for establishing a reverse index according to the corresponding strategy according to the size of the keyword set.
Further, the inverted index creating module specifically includes:
the size judgment module of the keyword set is used for judging the value of the size K of the keyword set;
a pinyin resource index establishing module, configured to, when K is less than or equal to 2, obtain a pronunciation of each word in the keyword set, obtain a pronunciation set { a1, a 2.., an }, and establish a corresponding pinyin resource index for each pronunciation in the pronunciation set;
and the character resource index establishing module is used for acquiring the code of each character in the keyword set when K is more than 2 to obtain a character code set { c1, c 2.. once, cn }, and establishing a corresponding character resource index for each character in the character code set.
Further, the voice search and result output module specifically includes:
the core keyword segmentation module is used for segmenting the obtained core keywords to obtain a user search keyword set;
the size judgment module of the user search keyword set is used for judging the value of the size M of the user search keyword set;
the first voice search result module is used for searching the pinyin resource index when M is less than or equal to 2 to obtain a primary search result; sequencing the initial search results according to the popularity of the resources to obtain a final voice search result;
the second voice search result module is used for searching the character resource index when M is larger than 2, and adopting different fault-tolerant strategies according to the value length of M to obtain a primary search result; and sequencing the initial search results according to the popularity of the resources to obtain a final voice search result.
Compared with the prior art, the invention pre-establishes the reverse index of the resources in the computing center and simultaneously updates the heat of the resources; collecting user voice data by using equipment; firstly, recognizing the user voice data and converting the user voice data into text data; performing semantic analysis on the text data to obtain core keywords; then, according to the length of the core keyword, different voice search strategies are adopted to obtain a primary search result; sequencing the initial search results according to the popularity of the resources to obtain a final voice search result; therefore, a new voice searching mode is provided, the search result can be quickly and accurately output within a certain error recognition range, and the voice searching experience effect of a user can be greatly improved.
Drawings
Fig. 1 is a flowchart of an implementation of a method for voice search according to an embodiment of the present invention.
Fig. 2 is a block diagram of a voice search apparatus according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The embodiment of the invention pre-establishes the reverse index of the resources in the computing center and updates the heat of the resources at the same time; collecting user voice data by using equipment; firstly, recognizing the user voice data and converting the user voice data into text data; performing semantic analysis on the text data to obtain core keywords; then, according to the length of the core keyword, different voice search strategies are adopted to obtain a primary search result; and the preliminary search results are sorted according to the popularity of the resources to obtain the final voice search result, so that the problems that the conventional voice recognition frequently makes mistakes, the search speed is low and the voice search experience of a user is poor are solved. The embodiment of the invention also provides a corresponding device, which is respectively explained in detail below.
Fig. 1 shows a flowchart of an implementation of a method for voice search according to an embodiment of the present invention.
In the embodiment of the invention, the voice searching method is applied to a computer, intelligent equipment, Al equipment, a robot and other human-computer interaction terminals needing voice input.
Referring to fig. 1, the method of voice search includes:
in step S1, the reverse index of the resource is established in advance, while the heat of the resource is updated.
In the embodiment of the invention, the preprocessing is needed in the computing center, the purpose of the preprocessing is to establish all resource inverted indexes in advance, which is convenient for fast positioning of resources, and the preprocessing is generally completed by the following 4 steps:
s101, numbering all resources to be searched and generating unique IDs of resource indexes.
S102, extracting keywords and sentences used for searching the resources according to the resource content.
S103, segmenting words, and segmenting the keywords and the sentences into single Chinese characters to obtain a keyword set.
And S104, establishing an inverted index according to a corresponding strategy according to the size of the keyword set. Wherein, inverted index (inverted index): also known as inverted index, posting archive or inverted archive, is an indexing method used to store a mapping of where a word is stored in a document or set of documents under a full-text search.
In specific implementation, assuming that the size of the keyword set is K, the step S104 further includes:
judging the value of the size K of the keyword set;
when K is less than or equal to 2, acquiring the pronunciation of each word in the keyword set to obtain a pronunciation set { a1, a 2.. an }, and establishing a corresponding pinyin resource index for each pronunciation in the pronunciation set;
and when K is larger than 2, acquiring the code of each word in the keyword set to obtain a character code set { c1, c 2.. once, cn }, and establishing a corresponding character resource index for each character in the character code set.
Through the mode, the reverse indexes of all resources, including the pinyin resource index and the character resource index, are pre-established in the computing center, and the voice search resources can be more conveniently and quickly positioned compared with the traditional technical scheme because the resources are searched by combining the characters and the pinyin.
And S2, collecting the voice data of the user. The voice of the user is collected through voice equipment of a human-computer interaction terminal such as a computer, intelligent equipment, Al equipment and a robot.
And S3, recognizing the user voice data and converting the user voice data into text data. The collected user voice is identified, namely automatic voice identification and voice-to-text identification, and the aim is to automatically convert the voice content of human into corresponding characters by a computing center of intelligent terminal equipment such as a computer and the like. Unlike speaker recognition and speaker verification, the latter attempts to recognize or verify the speaker who uttered the speech rather than the vocabulary content contained therein.
And S4, performing semantic analysis on the text data to obtain core keywords. That is, the intelligent terminal device performs semantic analysis on the text, which means that a long string of characters or contents is analyzed to obtain the abstract and the idea of the paragraph. Even further, the text of the entire article is organized. Thereby judging the user intention and outputting the user search core keyword.
S5, obtaining a preliminary search result by adopting different voice search strategies according to the length of the core keyword; and sequencing the initial search results according to the popularity of the resources to obtain a final voice search result.
When the invention is implemented, the character type resource index carries out different fault-tolerant strategies according to the length of the core key word. Specifically, the step S5 specifically includes the following steps:
s501, segmenting the obtained core keywords to obtain a user search keyword set.
S502, judging the value of the size M of the user search keyword set.
S503, when M is less than or equal to 2, searching the pinyin resource index to obtain a primary search result; and sequencing the initial search results according to the popularity of the resources to obtain a final voice search result.
Specifically, when M is less than or equal to 2, the method comprises the following steps:
s5031, the pronunciation of each keyword is obtained;
s5032, searching a resource set corresponding to each pronunciation according to the pinyin resource index;
s5033, calculating an intersection of the resource sets corresponding to the pronunciations;
s5034, judging whether the intersection is an empty set;
s5035, if the intersection is an empty set, the intersection is switched to the step S5041 for character resource search; if not, turning to S5036;
s5036, if the intersection is not an empty set, ordering the resources in the intersection and the resource heat updated according to the calculation center of S1, and obtaining a final voice search result.
In the following, M ═ 2 is exemplified:
firstly, acquiring the pronunciation of each character;
assuming that a1, a2, b1 and b2 are set, wherein a1 and a2 are the pronunciation of the first Chinese character, and b1 and b2 are the pronunciation of the second character, the resource sets U1 corresponding to a1 and a2 and U2 corresponding to b1 and b2 are obtained according to the pinyin resource indexes;
obtaining an intersection U0 of the U1 and U2 sets;
judging whether the set U0 is an empty set;
if the set U0 is an empty set, the process proceeds to step S5041 to perform a character resource search, and if the set U0 is not an empty set, the resources in the set U0 are ranked according to the resource heat updated by the calculation center of S1, and the final voice search result is input.
S504, when M is larger than 2, searching the character resource index, and adopting different fault-tolerant strategies according to the value length of M to obtain a primary search result; and sequencing the initial search results according to the popularity of the resources to obtain a final voice search result.
In specific implementation, when M is more than 2, the method is carried out according to the following steps:
s5041, acquiring character codes corresponding to the keywords;
s5042, searching and obtaining a resource set corresponding to each character according to the character resource index;
s5043, judging whether the value of M is less than or equal to 5; if yes, solving the intersection of the resource sets corresponding to each character, and turning to S5046; if not, the process proceeds to S5044. Specifically, when M is less than or equal to 5, a Chinese character is allowed to be lost in the intersection, that is, a resource appears in the above 4 resource sets, and then the resource is added into the intersection.
S5044, judging whether the value of M is less than or equal to 9; if yes, solving the intersection of the resource sets corresponding to each character, and turning to S5046; if not, the process proceeds to S5045. Specifically, when M ≦ 9, two kanji characters are allowed to be lost in the intersection, i.e., one resource appears in more than the above 7 resource sets, and is added to the intersection.
And when S5045 and M is greater than 9, intersection is calculated for the resource set corresponding to each character, and the process goes to S5046. Specifically, when M > 9, three kanji characters are allowed to be lost in the intersection, i.e., one resource appears in more than the above 6 resource sets, and is added to the intersection.
S5046, judging whether the intersection is an empty set or not, and if so, adopting a bottom-in-pocket strategy; if not, the step goes to S5047;
and S5047, sequencing the resources in the intersection according to the updated resource heat of the calculation center of S1, and obtaining a final voice search result.
In the following, M is set to 9:
firstly, acquiring character codes corresponding to each keyword to obtain c1, c2, a., c 9;
according to the character resource index, a resource set U1, U2, U9 corresponding to each character is searched;
intersection UN of the U1, U2, U9 sets is found. Since the rule K is 9, two chinese characters are allowed to be lost: a resource appears in more than 7 sets above and is added to the UN.
Judging whether the intersection UN is an empty set, if so, adopting a bottom-holding strategy, namely returning the resource with the highest heat degree in the resource set in the single character; if not, sequencing the resources in the intersection according to the updated resource heat of the calculation center of S1, and obtaining the final voice search result.
In summary, the voice search scheme provided by the embodiment of the present invention combines the characters and the pinyin to perform resource search, and compared with the conventional scheme, the voice search scheme provided by the embodiment of the present invention has two advantages:
1. compared with simple character search, the search method provided by the text adopts a pinyin search mode when the user search keyword set is less than 2, and can solve the problems of inaccurate pronunciation, dialect error and the like in short input; when the search keyword set of the user is more than 2, a mode of combining fault tolerance and character matching is adopted, the input keyword information of the user can be effectively utilized, and the search speed is not influenced while the identification error is covered.
2. Compared with simple pronunciation search, the search method provided by the text adopts a character and fault-tolerant combined mode to search when the length of the search keyword of the user is more than 2, and makes full use of the input information of the user, so that the search space is greatly reduced, and the search is more accurate and rapid.
It should be understood that, in the above embodiments, the order of execution of the steps is not meant to imply any order, and the order of execution of the steps should be determined by their function and inherent logic, and should not limit the implementation process of the embodiments of the present invention.
Fig. 2 is a block diagram illustrating a voice search apparatus according to an embodiment of the present invention, and only a part related to the embodiment of the present invention is shown for convenience of description.
In the embodiment of the present invention, the apparatus for implementing the method for voice search in the embodiment of fig. 1 may be a software unit, a hardware unit or a combination of software and hardware unit that is built in a computer or a server.
Referring to fig. 2, the apparatus includes:
the preprocessing module 21 is configured to pre-establish an inverted index of the resource and update the heat of the resource;
a voice data acquisition module 22 for acquiring user voice data;
the voice recognition module 23 is configured to recognize the user voice data and convert the user voice data into text data;
the semantic analysis module 24 is configured to perform semantic analysis on the text data to obtain core keywords;
the voice search and result output module 25 is used for obtaining a preliminary search result by adopting different voice search strategies according to the length of the core keyword; and sequencing the initial search results according to the popularity of the resources to obtain a final voice search result.
Further, the preprocessing module device further comprises:
the index ID generation module is used for numbering all resources to be searched and generating the unique ID of the resource index;
the keyword sentence extraction module is used for extracting keywords and sentences used for retrieving the resources according to the content of the resources;
the word segmentation module is used for segmenting words and the keywords and the sentences into single Chinese characters to obtain a keyword set;
and the reverse index establishing module is used for establishing a reverse index according to the corresponding strategy according to the size of the keyword set.
The reverse index establishing module specifically comprises:
the size judgment module of the keyword set is used for judging the value of the size K of the keyword set;
a pinyin resource index establishing module, configured to, when K is less than or equal to 2, obtain a pronunciation of each word in the keyword set, obtain a pronunciation set { a1, a 2.., an }, and establish a corresponding pinyin resource index for each pronunciation in the pronunciation set;
and the character resource index establishing module is used for acquiring the code of each character in the keyword set when K is more than 2 to obtain a character code set { c1, c 2.. once, cn }, and establishing a corresponding character resource index for each character in the character code set.
The voice search and result output module specifically comprises:
the core keyword segmentation module is used for segmenting the obtained core keywords to obtain a user search keyword set;
the size judgment module of the user search keyword set is used for judging the value of the size M of the user search keyword set;
the first voice search result module is used for searching the pinyin resource index when M is less than or equal to 2 to obtain a primary search result; sequencing the initial search results according to the popularity of the resources to obtain a final voice search result;
the second voice search result module is used for searching the character resource index when M is larger than 2, and adopting different fault-tolerant strategies according to the value length of M to obtain a primary search result; and sequencing the initial search results according to the popularity of the resources to obtain a final voice search result.
It should be noted that the apparatus in the embodiment of the present invention may be configured to implement all technical solutions in the foregoing method embodiments, and the functions of each functional module may be implemented specifically according to the method in the foregoing method embodiments, and the specific implementation process may refer to the relevant description in the foregoing example, which is not described herein again.
In summary, the embodiment of the present invention pre-establishes the reverse index of the resource in the computing center, and updates the heat of the resource at the same time; collecting user voice data by using equipment; firstly, recognizing the user voice data and converting the user voice data into text data; performing semantic analysis on the text data to obtain core keywords; then, according to the length of the core keyword, different voice search strategies are adopted to obtain a primary search result; sequencing the initial search results according to the popularity of the resources to obtain a final voice search result; therefore, a new voice searching mode is provided, the search result can be quickly and accurately output within a certain error recognition range, and the voice searching experience effect of a user can be greatly improved.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the embodiments provided in the present application, it should be understood that the disclosed method and apparatus can be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules or units is only one logical division, and there may be other divisions when actually implemented, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units and modules in the embodiments of the present invention may be integrated into one processing unit, or each unit and module may exist alone physically, or two or more units and modules may be integrated into one unit.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (4)
1. A method of voice searching, the method comprising:
establishing an inverted index of the resource in advance, and updating the heat of the resource;
collecting user voice data;
recognizing the user voice data and converting the user voice data into text data;
performing semantic analysis on the text data to obtain core keywords;
obtaining a preliminary search result by adopting different voice search strategies according to the length of the core keyword; and the preliminary search results are sorted according to the popularity of the resources to obtain the final voice search result,
the pre-established reverse index of the resource specifically includes:
numbering all resources to be searched and generating a unique ID of a resource index;
extracting key words and sentences used for retrieving the resources according to the content of the resources;
performing word segmentation, and segmenting the keywords and the sentences into single Chinese characters to obtain a keyword set;
establishing an inverted index according to a corresponding strategy according to the size of the keyword set;
the establishing of the inverted index according to the corresponding strategy according to the size of the keyword set specifically comprises:
judging the value of the size K of the keyword set;
when K is less than or equal to 2, acquiring the pronunciation of each word in the keyword set to obtain a pronunciation set { a1, a 2.. an }, and establishing a corresponding pinyin resource index for each pronunciation in the pronunciation set;
and when K is larger than 2, acquiring the code of each word in the keyword set to obtain a character code set { c1, c 2.. once, cn }, and establishing a corresponding character resource index for each character in the character code set.
2. The method of voice search according to claim 1, wherein different voice search strategies are adopted to obtain a preliminary search result according to the length of the core keyword; and ranking the initial search results according to the popularity of the resources to obtain a final voice search result, which specifically comprises the following steps:
segmenting the obtained core keywords to obtain a user search keyword set;
judging the value of the size M of the user search keyword set;
when M is less than or equal to 2, searching the pinyin resource index to obtain a primary search result; sequencing the initial search results according to the popularity of the resources to obtain a final voice search result;
when M is larger than 2, searching the character resource index, and adopting different fault-tolerant strategies according to the value length of M to obtain a primary search result; and sequencing the initial search results according to the popularity of the resources to obtain a final voice search result.
3. An apparatus for voice searching, the apparatus comprising:
the preprocessing module is used for pre-establishing an inverted index of the resource and updating the heat of the resource;
the voice data acquisition module is used for acquiring user voice data;
the voice recognition module is used for recognizing the user voice data and converting the user voice data into text data;
the semantic analysis module is used for performing semantic analysis on the text data to obtain core keywords;
the voice search and result output module is used for obtaining a preliminary search result by adopting different voice search strategies according to the length of the core keyword; sequencing the initial search results according to the popularity of the resources to obtain a final voice search result;
the preprocessing module further comprises:
the index ID generation module is used for numbering all resources to be searched and generating the unique ID of the resource index;
the keyword sentence extraction module is used for extracting keywords and sentences used for retrieving the resources according to the content of the resources;
the word segmentation module is used for segmenting words and the keywords and the sentences into single Chinese characters to obtain a keyword set;
the reverse index establishing module is used for establishing a reverse index according to the size of the keyword set and a corresponding strategy;
the reverse index establishing module specifically comprises:
the size judgment module of the keyword set is used for judging the value of the size K of the keyword set;
a pinyin resource index establishing module, configured to, when K is less than or equal to 2, obtain a pronunciation of each word in the keyword set, obtain a pronunciation set { a1, a 2.., an }, and establish a corresponding pinyin resource index for each pronunciation in the pronunciation set;
and the character resource index establishing module is used for acquiring the code of each character in the keyword set when K is more than 2 to obtain a character code set { c1, c 2.. once, cn }, and establishing a corresponding character resource index for each character in the character code set.
4. The apparatus for voice search according to claim 3, wherein the voice search and result output module comprises:
the core keyword segmentation module is used for segmenting the obtained core keywords to obtain a user search keyword set;
the size judgment module of the user search keyword set is used for judging the value of the size M of the user search keyword set;
the first voice search result module is used for searching the pinyin resource index when M is less than or equal to 2 to obtain a primary search result; sequencing the initial search results according to the popularity of the resources to obtain a final voice search result;
the second voice search result module is used for searching the character resource index when M is larger than 2, and adopting different fault-tolerant strategies according to the value length of M to obtain a primary search result; and sequencing the initial search results according to the popularity of the resources to obtain a final voice search result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811645648.9A CN109727591B (en) | 2018-12-29 | 2018-12-29 | Voice search method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811645648.9A CN109727591B (en) | 2018-12-29 | 2018-12-29 | Voice search method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109727591A CN109727591A (en) | 2019-05-07 |
| CN109727591B true CN109727591B (en) | 2020-08-11 |
Family
ID=66299402
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811645648.9A Active CN109727591B (en) | 2018-12-29 | 2018-12-29 | Voice search method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109727591B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113421565A (en) * | 2021-07-19 | 2021-09-21 | 北京百度网讯科技有限公司 | Search method, search device, electronic equipment and storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105590626A (en) * | 2015-12-29 | 2016-05-18 | 百度在线网络技术（北京）有限公司 | Continuous speech man-machine interaction method and system |
| JP2017107472A (en) * | 2015-12-11 | 2017-06-15 | 日本電信電話株式会社 | Character string search method and device |
-
2018
- 2018-12-29 CN CN201811645648.9A patent/CN109727591B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2017107472A (en) * | 2015-12-11 | 2017-06-15 | 日本電信電話株式会社 | Character string search method and device |
| CN105590626A (en) * | 2015-12-29 | 2016-05-18 | 百度在线网络技术（北京）有限公司 | Continuous speech man-machine interaction method and system |
Non-Patent Citations (1)
| Title |
|---|
| 云环境下个性化推送搜索引擎的设计;苟廷熹;《中国优秀硕士学位论文全文数据库信息科技辑》;20120815;第I138卷(第2012/08期);第3页第3段，第19页第2段，第20页第1，3-4段，第22页第1段，第23页第6段，第25页第2-6段,第26页6-7段 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109727591A (en) | 2019-05-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109408526B (en) | SQL sentence generation method, device, computer equipment and storage medium | |
| CN109918680B (en) | Entity identification method and device and computer equipment | |
| CN108287858B (en) | Semantic extraction method and device for natural language | |
| CN107291783B (en) | Semantic matching method and intelligent equipment | |
| CN106776544B (en) | Character relation recognition method and device and word segmentation method | |
| CN104142915B (en) | A kind of method and system adding punctuate | |
| CN110020424B (en) | Contract information extraction method and device and text information extraction method | |
| CN109147767B (en) | Method, device, computer equipment and storage medium for recognizing numbers in voice | |
| CN111241237B (en) | Intelligent question-answer data processing method and device based on operation and maintenance service | |
| CN103365925B (en) | Obtain polyphone phonetic, method based on phonetic retrieval and related device thereof | |
| CN110020422A (en) | The determination method, apparatus and server of Feature Words | |
| CN107330011A (en) | The recognition methods of the name entity of many strategy fusions and device | |
| CN106708929B (en) | Video program searching method and device | |
| CN106570180A (en) | Artificial intelligence based voice searching method and device | |
| CN103971684A (en) | Method and system for adding punctuations and method and device for establishing language model for adding punctuations | |
| CN111310440A (en) | Text error correction method, device and system | |
| CN110929520B (en) | Unnamed entity object extraction method and device, electronic equipment and storage medium | |
| CN112214576B (en) | Public opinion analysis method, public opinion analysis device, terminal equipment and computer readable storage medium | |
| CN111488468A (en) | Geographic information knowledge point extraction method and device, storage medium and computer equipment | |
| CN110808032A (en) | Voice recognition method and device, computer equipment and storage medium | |
| CN109299233A (en) | Text data processing method, device, computer equipment and storage medium | |
| CN112182145A (en) | Text similarity determination method, device, equipment and storage medium | |
| CN110309504B (en) | Text processing method, device, equipment and storage medium based on word segmentation | |
| CN106570196B (en) | Video program searching method and device | |
| CN110874408B (en) | Model training method, text recognition device and computing equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |