CN105389307A - Statement intention category identification method and apparatus - Google Patents
Statement intention category identification method and apparatus Download PDFInfo
- Publication number
- CN105389307A CN105389307A CN201510874529.0A CN201510874529A CN105389307A CN 105389307 A CN105389307 A CN 105389307A CN 201510874529 A CN201510874529 A CN 201510874529A CN 105389307 A CN105389307 A CN 105389307A
- Authority
- CN
- China
- Prior art keywords
- question sentence
- keyword
- dictionary
- question
- word
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/247—Thesauruses; Synonyms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Machine Translation (AREA)
Abstract
The present invention discloses a statement intention category identification method and apparatus. The method comprises: providing question and answer log information, wherein each piece of question and answer log information comprises: a question and an intention category; according to keywords obtained from the questions, generating a dictionary; according to the dictionary, performing vectorizing calculation on each question to obtain a vector of each question; according to the vector of each question and corresponding intention category, performing classification model training to obtain an intention classification model; and performing vectorizing calculation on a current question to obtain a vector of the current question, and according to the vector of the current question and the intention classification model, determining the intention category corresponding to the current question. By virtue of the technical scheme provided by the present invention, a question intention of a user can be identified preliminarily and accurately, so that the execution efficiency of semantic understanding is improved, the cost is reduced, the question response time is shortened, and the user experience is improved.
Description
Technical field
The present invention relates to field of computer technology, particularly relate to the intention classification recognition methods of a kind of statement and device.
Background technology
In the prior art, when carrying out intention to the question sentence of user and analyzing, be directly all template question sentences in current question sentence and knowledge base are carried out Similarity Measure, a formwork structure the highest according to similarity, analyze the intention of user, understand question sentence semanteme, arrange and return corresponding problem answers.But above-mentioned process, owing to will calculate similarity with all template question sentences for each file, make calculated amount very large, thus it is long to result in user's question sentence intention analytical calculation time, the problem that counting yield is low.
Summary of the invention
In view of the above problems, the present invention is proposed to provide a kind of statement intention classification recognition methods and device overcoming the problems referred to above or solve the problem at least in part.
The invention provides the recognition methods of a kind of statement intention classification, comprising:
There is provided Question and Answer log information, every bar Question and Answer log information comprises: question sentence and intention classification;
Dictionary is generated according to the keyword obtained from question sentence;
Respectively vectorization calculating is carried out to each question sentence according to dictionary, obtain the vector of each question sentence;
Carry out the training of disaggregated model according to the vector of each question sentence and the intention classification of correspondence, obtain intent classifier model;
Vectorization calculating is carried out to current question sentence, obtains the vector of current question sentence, determine according to the vector sum intent classifier model of current question sentence the intention classification that current question sentence is corresponding.
Present invention also offers a kind of statement intention classification recognition device, comprising:
There is provided module, for providing Question and Answer log information, every bar Question and Answer log information comprises: question sentence and intention classification;
Generation module, for generating dictionary according to the keyword obtained from question sentence;
Computing module, for carrying out vectorization calculating to each question sentence respectively according to dictionary, obtains the vector of each question sentence;
Training module, for carrying out the training of disaggregated model according to the vector of each question sentence and the intention classification of correspondence, obtains intent classifier model;
Identification module, for carrying out vectorization calculating to current question sentence, obtains the vector of current question sentence, determines according to the vector sum intent classifier model of current question sentence the intention classification that current question sentence is corresponding.
Beneficial effect of the present invention is as follows:
By obtaining keyword in the question sentence of log information from question answering system, and utilize these keywords tentatively to determine the intention of user, solve user's question sentence intention analytical calculation time in prior art long, the problem that counting yield is low, the question sentence intention of user can be identified tentatively exactly, improve the execution efficiency of semantic understanding, reduce cost, save the time that answer is replied, improve the experience of user.
Above-mentioned explanation is only the general introduction of technical solution of the present invention, in order to technological means of the present invention can be better understood, and can be implemented according to the content of instructions, and can become apparent, below especially exemplified by the specific embodiment of the present invention to allow above and other objects of the present invention, feature and advantage.
Accompanying drawing explanation
By reading hereafter detailed description of the preferred embodiment, various other advantage and benefit will become cheer and bright for those of ordinary skill in the art.Accompanying drawing only for illustrating the object of preferred implementation, and does not think limitation of the present invention.And in whole accompanying drawing, represent identical parts by identical reference symbol.In the accompanying drawings:
Fig. 1 is the process flow diagram of the statement intention classification recognition methods of the embodiment of the present invention;
Fig. 2 is the process flow diagram of the detailed process of the statement intention classification recognition methods of the embodiment of the present invention;
Fig. 3 is the structural representation of the statement intention classification recognition device of the embodiment of the present invention.
Embodiment
Below with reference to accompanying drawings exemplary embodiment of the present disclosure is described in more detail.Although show exemplary embodiment of the present disclosure in accompanying drawing, however should be appreciated that can realize the disclosure in a variety of manners and not should limit by the embodiment set forth here.On the contrary, provide these embodiments to be in order to more thoroughly the disclosure can be understood, and complete for the scope of the present disclosure can be conveyed to those skilled in the art.
Long in order to solve the prior art user question sentence intention analytical calculation time, the problem that counting yield is low, the invention provides the intention classification recognition methods of a kind of statement and device, below in conjunction with accompanying drawing and embodiment, is further elaborated to the present invention.Should be appreciated that specific embodiment described herein only in order to explain the present invention, do not limit the present invention.
Embodiment of the method
According to embodiments of the invention, provide the recognition methods of a kind of statement intention classification, Fig. 1 is the process flow diagram of the statement intention classification recognition methods of the embodiment of the present invention, and as shown in Figure 1, the statement intention classification recognition methods according to the embodiment of the present invention comprises following process:
Step 101, provides Question and Answer log information, and every bar Question and Answer log information comprises: question sentence and intention classification.Wherein, above-mentioned intention classification can comprise: weather, shopping, work, tourism etc.;
Step 102, generates dictionary according to the keyword obtained from question sentence.
Generate dictionary in step 102 and specifically comprise following process:
Step 1021, generates dictionary according to the keyword obtained from question sentence and comprises:
Step 1022, carries out pre-service to the language material obtained from question sentence, obtains text data.Wherein, pre-service comprises: be text formatting by the uniform format of question sentence, filters one or more in dirty word, sensitive word and stop words, and the text data after filtering is divided into multirow according to punctuate.Such as, above-mentioned punctuate can be question mark, exclamation, branch or fullstop, that is, the text data after filtration can be divided into multirow according to question mark, exclamation, branch or fullstop.
Step 1023, carries out word segmentation processing to text data, obtains multiple language material word.In embodiments of the present invention, word segmentation processing can adopt one or more in the two-way maximum matching method of dictionary, viterbi method, HMM method and CRF method.
Step 1024, carries out filtration treatment to language material word, obtains the dictionary comprising multiple keyword.Wherein, filtration treatment comprises any one or two kinds of modes following:
Mode one: filter language material word according to part of speech, retains noun, verb and adjective;
Mode two: filter language material word according to the frequency, retain the language material word that the frequency is greater than frequency threshold value, wherein, the frequency refers to the frequency that language material word occurs in corpus data or number of times.
In the present embodiment, first according to part of speech, language material word is filtered, only retain noun, verb and adjective, remove the language material word of other part of speech; Then according to the frequency, the noun retained, verb and adjective are filtered, retain the language material word that the frequency is greater than frequency threshold value, thus for the frequency is greater than the noun of frequency threshold value, verb and adjective in dictionary.
In other embodiments of the invention, only can filter according to part of speech, also can only filter according to the frequency, can also first carry out filtering filtering according to part of speech according to the frequency, it be all within protection scope of the present invention again.
Step 1025, carries out dimension-reduction treatment to dictionary.
Wherein, dimension-reduction treatment specifically comprises: the intention classification corresponding according to Question and Answer log statistics question sentence, and calculate the information entropy of each keyword in dictionary, keyword information entropy being less than information entropy threshold value is deleted from dictionary, wherein, information entropy represents the probability that this keyword occurs in each intention classification.The information entropy calculating keyword in dictionary comprises: calculate the probability that in dictionary, each keyword occurs in each intention classification.
The computing formula of information entropy is: H (X)=-Σ p (x
i) logp (x
i), wherein, H (X) represents the information entropy of keyword, p (x
i) represent the probability that keyword occurs in i-th intention classification, i=1,2 ..., n, n are the number of intention classification.
Because the information entropy of keyword can directly be calculated according to above-mentioned computing formula, computation complexity is very low, therefore, the keyword that the technical scheme of the embodiment of the present invention utilizes information entropy quick and precisely information entropy can be less than information entropy threshold value is deleted from dictionary, substantially increases accuracy rate and the efficiency of dictionary dimensionality reduction.
It should be noted that, in other embodiments of the invention, other existing method also can be adopted to carry out dimension-reduction treatment, as: SVD, LDA or PCA etc., it does not affect protection scope of the present invention.
The embodiment of the present invention, by carrying out dimensionality reduction to dictionary, makes dictionary more accurately and simplify, follow-up carry out vectorization calculating and the disaggregated model training of question sentence time, greatly reduce the complexity of calculating, effectively improve counting yield.
Step 103, carries out vectorization calculating to each question sentence respectively according to dictionary, obtains the vector of each question sentence.
In step 103, carry out vectorization calculating to question sentence to comprise:
Step 1031, carries out pre-service and word segmentation processing to question sentence.
Wherein, pre-service specifically comprises: be text formatting by the uniform format of question sentence, filters one or more in dirty word, sensitive word and stop words, and the text data after filtering is divided into multirow according to punctuate.Such as, above-mentioned punctuate can be question mark, exclamation, branch or fullstop, that is, the text data after filtration can be divided into multirow according to question mark, exclamation, branch or fullstop.
Described word segmentation processing can adopt in the two-way maximum matching method of dictionary, viterbi method, HMM method and CRF method one or more.
Step 1032, according to the TF-IDF value of the keyword occurred in the size of dictionary and corresponding question sentence, vector is generated to this question sentence, wherein, the dimension of vector is the size of dictionary, the value of each dimension is: the dimension values not appearing at the word in dictionary in corresponding question sentence is 0, and the dimension values appearing at the keyword in dictionary in corresponding question sentence is the TF-IDF value of this keyword.
Wherein, the TF-IDF value of keyword obtains in the following manner:
The business obtained, divided by the number of question sentence comprising keyword, takes the logarithm and obtains the IDF value of keyword by the question sentence total number 1, comprised by Question and Answer log;
2, calculate the frequency that keyword occurs in corresponding question sentence, determine TF value;
3, TF value is multiplied by the TF-IDF value that IDF value obtains keyword.
Step 104, carries out the training of disaggregated model according to the vector of each question sentence and the intention classification of correspondence, obtain intent classifier model.
In embodiments of the present invention, the method for carrying out disaggregated model training can comprise: one or both in SVM and naive Bayesian.
Step 105, carries out vectorization calculating to current question sentence, obtains the vector of current question sentence, determines according to the vector sum intent classifier model of current question sentence the intention classification that current question sentence is corresponding.
In step 105, can perform with reference to step 1031-1032 the process that current question sentence carries out vectorization calculating.
Below in conjunction with accompanying drawing, the technique scheme of the embodiment of the present invention is described in detail.
Fig. 2 is the process flow diagram of the detailed process of the statement intention classification recognition methods of the embodiment of the present invention, as shown in Figure 2, specifically comprises following process:
Step 201, obtains the question sentence in Question and Answer log information and intention classification, and obtains the corresponding relation between question sentence and intention classification, is stored into intention classification table;
Step 202, unified to the language material obtained from question sentence is text formatting, and filters invalid form, removes one or more in dirty word, sensitive word and stop words; Text data after filtering is split by large punctuate preservation of embarking on journey, and wherein, large punctuate can be question mark, exclamation, branch or fullstop;
Step 203, one or more in the two-way maximum matching method of employing dictionary, viterbi method, HMM method and CRF method carry out word segmentation processing to the text data after branch, wherein, by space-separated between each word, such as: w1w2w3w4w5 ...
Step 204, carries out filtration treatment to language material word, obtains the dictionary comprising multiple keyword; Wherein, filtration treatment adopts any one or two kinds of modes following: mode one: filter language material word according to part of speech, retains noun, verb and adjective; Mode two: filter language material word according to the frequency, retain the language material word that the frequency is greater than frequency threshold value, wherein, the frequency refers to the frequency that language material word occurs in corpus data or number of times.
Step 205, add up the number of times that each keyword occurs in difference intention classification, and calculate the information entropy of each keyword in dictionary, wherein, information entropy represents the probability that this keyword occurs in each intention classification, the information entropy calculated is saved in information entropy database, the information entropy of keyword is obtained from information entropy database, judge whether the information entropy of keyword is less than the information entropy threshold value pre-set, and keyword information entropy being greater than information entropy threshold value is retained in dictionary, dictionary is kept in keyword dictionary database.
Step 206, the business obtained, divided by the number of question sentence comprising keyword, takes the logarithm and obtains the IDF value of keyword by the question sentence total number comprised by Question and Answer log; Calculate the frequency that keyword occurs in corresponding question sentence, determine TF value; TF value is multiplied by the TF-IDF value that IDF value obtains keyword.The TF-IDF value of keyword is stored into TF-IDF property data base.
Particularly, the formula calculating the IDF value of keyword is as follows: wherein, and D represents question sentence sum, { j:t
i∈ d
jrepresent the number comprising the question sentence of keyword.
Step 207, according to the TF-IDF value of the keyword occurred in the size of dictionary and corresponding question sentence, vector is generated to this question sentence, wherein, the dimension of vector is the size of dictionary, the value of each dimension is: the dimension values not appearing at the word in dictionary in corresponding question sentence is 0, and the dimension values appearing at the keyword in dictionary in corresponding question sentence is the TF-IDF value of this keyword.
Step 208, carries out the training of disaggregated model according to the vector of each question sentence and the intention classification of correspondence, obtain intent classifier model, the method for carrying out disaggregated model training comprises: one or both in SVM and naive Bayesian.In embodiments of the present invention, employing is svm classifier training pattern.In other embodiments, can also adopt naive Bayesian or, adopt in SVM and naive Bayesian to combine and carry out disaggregated model training.
Step 209, vectorization calculating is carried out to current question sentence, particularly, first, pre-service and participle are carried out to question sentence, according to the TF-IDF value being stored in TF-IDF property data base corresponding to the keyword occurred in the size of the dictionary in keyword dictionary database and corresponding question sentence, vector is generated to this question sentence, obtain the vector of current question sentence, determine according to the vector sum intent classifier model of current question sentence the intention classification that current question sentence is corresponding.
In sum, by obtaining keyword in the question sentence of log information from question answering system, and utilize these keywords tentatively to determine the intention of user, solve user's question sentence intention analytical calculation time in prior art long, the problem that counting yield is low, the question sentence intention of user can be identified tentatively exactly, improve the execution efficiency of semantic understanding, reduce cost, saved the time that answer is replied, improve the experience of user.
Device embodiment
According to embodiments of the invention, provide a kind of statement intention classification recognition device, Fig. 3 is the structural representation of the statement intention classification recognition device of the embodiment of the present invention, as shown in Figure 3, statement intention classification recognition device according to the embodiment of the present invention comprises: provide module 30, generation module 31, computing module 32, training module 33 and identification module 34, be described in detail below to the modules of the embodiment of the present invention.
There is provided module 30, for providing Question and Answer log information, every bar Question and Answer log information comprises: question sentence and intention classification;
Generation module 31, for generating dictionary according to the keyword obtained from question sentence; Generation module 31 specifically comprises:
Pre-service submodule, for carrying out pre-service to the language material obtained from question sentence, obtains text data; Particularly, the uniform format of question sentence is text formatting by pre-service submodule, filters one or more in dirty word, sensitive word and stop words, and the text data after filtering is divided into multirow according to punctuate.
Word segmentation processing submodule, for carrying out word segmentation processing to text data, obtains multiple language material word; Word segmentation processing adopt in dictionary two-way maximum matching method, viterbi method, HMM method and CRF method one or more.
Filtration treatment submodule, for carrying out filtration treatment to language material word, obtains the dictionary comprising multiple keyword; Filtration treatment submodule specifically for: adopt any one or two kinds of modes following to carry out filtration treatment: mode one: filter language material word according to part of speech, retain noun, verb and adjective; Mode two: filter language material word according to the frequency, retains the language material word that the frequency is greater than frequency threshold value.
Dimensionality reduction submodule, for carrying out dimension-reduction treatment to dictionary.Dimensionality reduction submodule specifically for: according to intention classification corresponding to Question and Answer log statistics question sentence, calculate the information entropy of each keyword in dictionary, keyword information entropy being less than information entropy threshold value is deleted from dictionary, and wherein, information entropy represents the probability that this keyword occurs in each intention classification.
Because the information entropy of keyword can directly be calculated according to above-mentioned computing formula, computation complexity is very low, therefore, the keyword that the technical scheme of the embodiment of the present invention utilizes information entropy quick and precisely information entropy can be less than information entropy threshold value is deleted from dictionary, substantially increases accuracy rate and the efficiency of dictionary dimensionality reduction.
The embodiment of the present invention, by carrying out dimensionality reduction to dictionary, makes dictionary more accurately and simplify, follow-up carry out vectorization calculating and the disaggregated model training of question sentence time, greatly reduce the complexity of calculating, effectively improve counting yield.
Computing module 32, for carrying out vectorization calculating to each question sentence respectively according to dictionary, obtains the vector of each question sentence; Computing module 32 specifically for:
Pre-service and participle are carried out to question sentence; Pre-service specifically comprises: be text formatting by the uniform format of question sentence, filters one or more in dirty word, sensitive word and stop words, and the text data after filtering is divided into multirow according to punctuate.
According to the TF-IDF value of the keyword occurred in the size of dictionary and corresponding question sentence, vector is generated to this question sentence, wherein, the dimension of vector is the size of dictionary, the value of each dimension is: the dimension values not appearing at the word in dictionary in corresponding question sentence is 0, and the dimension values appearing at the keyword in dictionary in corresponding question sentence is the TF-IDF value of this keyword.Wherein, the TF-IDF value calculating keyword is:
The business obtained, divided by the number of question sentence comprising keyword, takes the logarithm and obtains the IDF value of keyword by the question sentence total number comprised by Question and Answer log; Calculate the frequency that keyword occurs in corresponding question sentence, determine TF value; TF value is multiplied by the TF-IDF value that IDF value obtains keyword.
Training module 33, for carrying out the training of disaggregated model according to the vector of each question sentence and the intention classification of correspondence, obtains intent classifier model; The method that training module 33 carries out disaggregated model training comprises: one or both in SVM and naive Bayesian.
Identification module 34, for carrying out vectorization calculating to current question sentence, obtains the vector of current question sentence, determines according to the vector sum intent classifier model of current question sentence the intention classification that current question sentence is corresponding.
In sum, by obtaining keyword in the question sentence of log information from question answering system, and utilize these keywords tentatively to determine the intention of user, solve user's question sentence intention analytical calculation time in prior art long, the problem that counting yield is low, the question sentence intention of user can be identified tentatively exactly, improve the execution efficiency of semantic understanding, reduce cost, saved the time that answer is replied, improve the experience of user.
Obviously, those skilled in the art can carry out various change and modification to the present invention and not depart from the spirit and scope of the present invention.Like this, if these amendments of the present invention and modification belong within the scope of the claims in the present invention and equivalent technologies thereof, then the present invention is also intended to comprise these change and modification.
Intrinsic not relevant to any certain computer, virtual system or miscellaneous equipment with display at this algorithm provided.Various general-purpose system also can with use based on together with this teaching.According to description above, the structure constructed required by this type systematic is apparent.In addition, the present invention is not also for any certain programmed language.It should be understood that and various programming language can be utilized to realize content of the present invention described here, and the description done language-specific is above to disclose preferred forms of the present invention.
In instructions provided herein, describe a large amount of detail.But can understand, embodiments of the invention can be put into practice when not having these details.In some instances, be not shown specifically known method, structure and technology, so that not fuzzy understanding of this description.
Similarly, be to be understood that, in order to simplify the disclosure and to help to understand in each inventive aspect one or more, in the description above to exemplary embodiment of the present invention, each feature of the present invention is grouped together in single embodiment, figure or the description to it sometimes.But, the method for the disclosure should be construed to the following intention of reflection: namely the present invention for required protection requires feature more more than the feature clearly recorded in each claim.Or rather, as claims below reflect, all features of disclosed single embodiment before inventive aspect is to be less than.Therefore, the claims following embodiment are incorporated to this embodiment thus clearly, and wherein each claim itself is as independent embodiment of the present invention.
Those skilled in the art are appreciated that and adaptively can change the module in the client in embodiment and they are arranged in one or more clients different from this embodiment.Block combiner in embodiment can be become a module, and multiple submodule or subelement or sub-component can be put them in addition.Except at least some in such feature and/or process or unit be mutually repel except, any combination can be adopted to combine all processes of all features disclosed in this instructions (comprising adjoint claim, summary and accompanying drawing) and so disclosed any method or client or unit.Unless expressly stated otherwise, each feature disclosed in this instructions (comprising adjoint claim, summary and accompanying drawing) can by providing identical, alternative features that is equivalent or similar object replaces.
In addition, those skilled in the art can understand, although embodiments more described herein to comprise in other embodiment some included feature instead of further feature, the combination of the feature of different embodiment means and to be within scope of the present invention and to form different embodiments.Such as, in the following claims, the one of any of embodiment required for protection can use with arbitrary array mode.
All parts embodiment of the present invention with hardware implementing, or can realize with the software module run on one or more processor, or realizes with their combination.It will be understood by those of skill in the art that the some or all functions of some or all parts be loaded with in the client of sequence network address that microprocessor or digital signal processor (DSP) can be used in practice to realize according to the embodiment of the present invention.The present invention can also be embodied as part or all equipment for performing method as described herein or device program (such as, computer program and computer program).Realizing program of the present invention and can store on a computer-readable medium like this, or the form of one or more signal can be had.Such signal can be downloaded from internet website and obtain, or provides on carrier signal, or provides with any other form.
The present invention will be described instead of limit the invention to it should be noted above-described embodiment, and those skilled in the art can design alternative embodiment when not departing from the scope of claims.In the claims, any reference symbol between bracket should be configured to limitations on claims.Word " comprises " not to be got rid of existence and does not arrange element in the claims or step.Word "a" or "an" before being positioned at element is not got rid of and be there is multiple such element.The present invention can by means of including the hardware of some different elements and realizing by means of the computing machine of suitably programming.In the unit claim listing some devices, several in these devices can be carry out imbody by same hardware branch.Word first, second and third-class use do not represent any order.Can be title by these word explanations.
Claims (18)
1. a statement intention classification recognition methods, is characterized in that, comprising:
There is provided Question and Answer log information, every bar Question and Answer log information comprises: question sentence and intention classification;
Dictionary is generated according to the keyword obtained from described question sentence;
Respectively vectorization calculating is carried out to each question sentence according to described dictionary, obtain the vector of each question sentence;
Carry out the training of disaggregated model according to the vector of each question sentence and the intention classification of correspondence, obtain intent classifier model;
Carry out vectorization calculating to current question sentence, obtain the vector of current question sentence, according to the vector sum of described current question sentence, the intention classification that current question sentence is corresponding determined by intent classifier model.
2. the method for claim 1, is characterized in that, generates dictionary comprise according to the keyword obtained from described question sentence:
Pre-service is carried out to the language material obtained from question sentence, obtains text data;
Word segmentation processing is carried out to described text data, obtains multiple language material word;
Filtration treatment is carried out to described language material word, obtains the dictionary comprising multiple keyword;
Dimension-reduction treatment is carried out to described dictionary.
3. method as claimed in claim 2, it is characterized in that, described dimension-reduction treatment comprises: the intention classification corresponding according to described Question and Answer log statistics question sentence, calculate the information entropy of each keyword in described dictionary, keyword information entropy being less than information entropy threshold value is deleted from described dictionary, wherein, described information entropy represents the probability that this keyword occurs in each intention classification.
4. the method for claim 1, is characterized in that, carries out vectorization calculating comprise question sentence:
Pre-service and word segmentation processing are carried out to described question sentence;
According to the TF-IDF value of the keyword occurred in the size of described dictionary and corresponding question sentence, vector is generated to this question sentence, wherein, the dimension of described vector is the size of described dictionary, the value of each dimension is: the dimension values not appearing at the word in dictionary in corresponding question sentence is 0, and the dimension values appearing at the keyword in dictionary in corresponding question sentence is the TF-IDF value of this keyword.
5. method as claimed in claim 4, it is characterized in that, the TF-IDF value of described keyword obtains in the following manner:
The business obtained, divided by the number of question sentence comprising described keyword, takes the logarithm and obtains the IDF value of described keyword by the question sentence total number comprised by Question and Answer log;
Calculate the frequency that described keyword occurs in corresponding question sentence, determine TF value;
Described TF value is multiplied by the TF-IDF value that described IDF value obtains described keyword.
6. the method as described in claim 2 or 4, it is characterized in that, described pre-service comprises: be text formatting by the uniform format of question sentence, filters one or more in dirty word, sensitive word and stop words, and the text data after filtering is divided into multirow according to punctuate.
7. the method as described in claim 2 or 4, is characterized in that, described word segmentation processing adopt in the two-way maximum matching method of dictionary, viterbi method, HMM method and CRF method one or more.
8. method as claimed in claim 2, is characterized in that, described filtration treatment adopts any one or two kinds of modes following:
According to part of speech, described language material word is filtered, retain noun, verb and adjective;
According to the frequency, described language material word is filtered, retain the language material word that the frequency is greater than frequency threshold value.
9. the method for claim 1, is characterized in that, the method for carrying out disaggregated model training comprises: one or both in SVM and naive Bayesian.
10. a statement intention classification recognition device, is characterized in that, comprising:
There is provided module, for providing Question and Answer log information, every bar Question and Answer log information comprises: question sentence and intention classification;
Generation module, for generating dictionary according to the keyword obtained from described question sentence;
Computing module, for carrying out vectorization calculating to each question sentence respectively according to described dictionary, obtains the vector of each question sentence;
Training module, for carrying out the training of disaggregated model according to the vector of each question sentence and the intention classification of correspondence, obtains intent classifier model;
Identification module, for carrying out vectorization calculating to current question sentence, obtains the vector of current question sentence, and according to the vector sum of described current question sentence, the intention classification that current question sentence is corresponding determined by intent classifier model.
11. devices as claimed in claim 10, it is characterized in that, described generation module specifically comprises:
Pre-service submodule, for carrying out pre-service to the language material obtained from question sentence, obtains text data;
Word segmentation processing submodule, for carrying out word segmentation processing to described text data, obtains multiple language material word;
Filtration treatment submodule, for carrying out filtration treatment to described language material word, obtains the dictionary comprising multiple keyword;
Dimensionality reduction submodule, for carrying out dimension-reduction treatment to described dictionary.
12. devices as claimed in claim 11, it is characterized in that, described dimensionality reduction submodule is specifically for the intention classification corresponding according to described Question and Answer log statistics question sentence, calculate the information entropy of each keyword in described dictionary, keyword information entropy being less than information entropy threshold value is deleted from described dictionary, wherein, described information entropy represents the probability that this keyword occurs in each intention classification.
13. devices as claimed in claim 10, is characterized in that, described computing module specifically for:
Pre-service and word segmentation processing are carried out to described question sentence;
According to the TF-IDF value of the keyword occurred in the size of described dictionary and corresponding question sentence, vector is generated to this question sentence, wherein, the dimension of described vector is the size of described dictionary, the value of each dimension is: the dimension values not appearing at the word in dictionary in corresponding question sentence is 0, and the dimension values appearing at the keyword in dictionary in corresponding question sentence is the TF-IDF value of this keyword.
14. devices as claimed in claim 13, is characterized in that, described computing module specifically for:
The business obtained, divided by the number of question sentence comprising described keyword, takes the logarithm and obtains the IDF value of described keyword by the question sentence total number comprised by Question and Answer log;
Calculate the frequency that described keyword occurs in corresponding question sentence, determine TF value;
Described TF value is multiplied by the TF-IDF value that described IDF value obtains described keyword.
15. devices as described in claim 11 or 13, it is characterized in that, described pre-service specifically for: be text formatting by the uniform format of question sentence, filter one or more in dirty word, sensitive word and stop words, and the text data after filtering is divided into multirow according to punctuate.
16. devices as described in claim 11 or 13, is characterized in that, described word segmentation processing adopt in the two-way maximum matching method of dictionary, viterbi method, HMM method and CRF method one or more.
17. devices as claimed in claim 11, is characterized in that, described filtration treatment submodule specifically adopts any one or two kinds of modes following to carry out filtration treatment:
According to part of speech, described language material word is filtered, retain noun, verb and adjective;
According to the frequency, described language material word is filtered, retain the language material word that the frequency is greater than frequency threshold value.
18. devices as claimed in claim 10, is characterized in that, described training module adopts one or both technology in SVM and naive Bayesian.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510874529.0A CN105389307A (en) | 2015-12-02 | 2015-12-02 | Statement intention category identification method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510874529.0A CN105389307A (en) | 2015-12-02 | 2015-12-02 | Statement intention category identification method and apparatus |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105389307A true CN105389307A (en) | 2016-03-09 |
Family
ID=55421604
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510874529.0A Pending CN105389307A (en) | 2015-12-02 | 2015-12-02 | Statement intention category identification method and apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105389307A (en) |
Cited By (46)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105955965A (en) * | 2016-06-21 | 2016-09-21 | 上海智臻智能网络科技股份有限公司 | Question information processing method and device |
| CN105975460A (en) * | 2016-05-30 | 2016-09-28 | 上海智臻智能网络科技股份有限公司 | Question information processing method and device |
| CN106056154A (en) * | 2016-05-27 | 2016-10-26 | 大连楼兰科技股份有限公司 | Fault code recognition and classification method |
| CN106649694A (en) * | 2016-12-19 | 2017-05-10 | 北京云知声信息技术有限公司 | Method and device for identifying user's intention in voice interaction |
| CN106649696A (en) * | 2016-12-19 | 2017-05-10 | 北京云知声信息技术有限公司 | Information classification method and device |
| CN107122404A (en) * | 2017-03-22 | 2017-09-01 | 北京晓数聚传媒科技有限公司 | A kind of user view data extracting method and device |
| CN107153672A (en) * | 2017-03-22 | 2017-09-12 | 中国科学院自动化研究所 | User mutual intension recognizing method and system based on Speech Act Theory |
| CN107193865A (en) * | 2017-04-06 | 2017-09-22 | 上海奔影网络科技有限公司 | Natural language is intended to understanding method and device in man-machine interaction |
| CN107291685A (en) * | 2016-04-13 | 2017-10-24 | 北京大学 | Method for recognizing semantics and semantics recognition system |
| CN107329948A (en) * | 2017-05-23 | 2017-11-07 | 努比亚技术有限公司 | Sentence describes time of origin presumption method, equipment and the storage medium of event |
| CN107943791A (en) * | 2017-11-24 | 2018-04-20 | 北京奇虎科技有限公司 | A kind of recognition methods of refuse messages, device and mobile terminal |
| CN108121722A (en) * | 2016-11-28 | 2018-06-05 | 渡鸦科技（北京）有限责任公司 | The construction method and device of knowledge base |
| CN108153737A (en) * | 2017-12-30 | 2018-06-12 | 北京中关村科金技术有限公司 | A kind of method of semantic classification, system and dialog process system |
| CN108228559A (en) * | 2016-12-22 | 2018-06-29 | 苏宁云商集团股份有限公司 | A kind of human-computer interaction realization method and system for customer service |
| CN108268617A (en) * | 2018-01-05 | 2018-07-10 | 阿里巴巴集团控股有限公司 | User view determines method and device |
| CN108334496A (en) * | 2018-01-30 | 2018-07-27 | 中国科学院自动化研究所 | Human-computer dialogue understanding method and system and relevant device for specific area |
| CN108334495A (en) * | 2018-01-30 | 2018-07-27 | 国家计算机网络与信息安全管理中心 | Short text similarity calculating method and system |
| CN108446286A (en) * | 2017-02-16 | 2018-08-24 | 阿里巴巴集团控股有限公司 | A kind of generation method, device and the server of the answer of natural language question sentence |
| CN108920579A (en) * | 2018-06-22 | 2018-11-30 | 龙马智芯（珠海横琴）科技有限公司 | Classification method, device, terminal device and the storage medium of sentence |
| CN109033075A (en) * | 2018-06-29 | 2018-12-18 | 北京百度网讯科技有限公司 | It is intended to matched method, apparatus, storage medium and terminal device |
| CN109241261A (en) * | 2018-08-30 | 2019-01-18 | 武汉斗鱼网络科技有限公司 | User's intension recognizing method, device, mobile terminal and storage medium |
| CN109388695A (en) * | 2018-09-27 | 2019-02-26 | 深圳前海微众银行股份有限公司 | User's intension recognizing method, equipment and computer readable storage medium |
| CN109388705A (en) * | 2017-08-07 | 2019-02-26 | 芋头科技（杭州）有限公司 | A kind of text intent classifier method |
| CN109522393A (en) * | 2018-10-11 | 2019-03-26 | 平安科技（深圳）有限公司 | Intelligent answer method, apparatus, computer equipment and storage medium |
| CN109766414A (en) * | 2019-01-18 | 2019-05-17 | 广东小天才科技有限公司 | Intention identification method and system |
| CN109815492A (en) * | 2019-01-04 | 2019-05-28 | 平安科技（深圳）有限公司 | A kind of intension recognizing method based on identification model, identification equipment and medium |
| CN109815308A (en) * | 2017-10-31 | 2019-05-28 | 北京小度信息科技有限公司 | The determination of intention assessment model and retrieval intension recognizing method, device |
| CN109857841A (en) * | 2018-12-05 | 2019-06-07 | 厦门快商通信息技术有限公司 | A kind of FAQ question sentence Text similarity computing method and system |
| CN109933779A (en) * | 2017-12-18 | 2019-06-25 | 苏宁云商集团股份有限公司 | User's intension recognizing method and system |
| CN109949637A (en) * | 2019-03-13 | 2019-06-28 | 广东小天才科技有限公司 | Automatic answering method and device for objective questions |
| CN110232114A (en) * | 2019-05-06 | 2019-09-13 | 平安科技（深圳）有限公司 | Sentence intension recognizing method, device and computer readable storage medium |
| CN110313154A (en) * | 2017-02-14 | 2019-10-08 | 微软技术许可有限责任公司 | Intelligent assistant with the information discrimination based on intention |
| CN110457466A (en) * | 2019-06-28 | 2019-11-15 | 谭浩 | Generate method, computer readable storage medium and the terminal device of interview report |
| CN110619050A (en) * | 2018-06-20 | 2019-12-27 | 华为技术有限公司 | Intention recognition method and equipment |
| CN110659175A (en) * | 2018-06-30 | 2020-01-07 | 中兴通讯股份有限公司 | Log trunk extraction method, log trunk classification method, log trunk extraction equipment and log trunk storage medium |
| CN110909142A (en) * | 2019-11-20 | 2020-03-24 | 腾讯科技（深圳）有限公司 | Question and sentence processing method and device of question-answer model, electronic equipment and storage medium |
| CN111128156A (en) * | 2019-12-10 | 2020-05-08 | 上海雷盎云智能技术有限公司 | Intelligent household equipment voice control method and device based on model training |
| CN111159403A (en) * | 2019-12-27 | 2020-05-15 | 广东高乐教育科技有限公司 | Intelligent classroom perception method and system |
| CN111507085A (en) * | 2019-11-25 | 2020-08-07 | 江苏艾佳家居用品有限公司 | Sentence pattern recognition method |
| US11010601B2 (en) | 2017-02-14 | 2021-05-18 | Microsoft Technology Licensing, Llc | Intelligent assistant device communicating non-verbal cues |
| CN112989839A (en) * | 2019-12-18 | 2021-06-18 | 中国科学院声学研究所 | Keyword feature-based intent recognition method and system embedded in language model |
| US11100384B2 (en) | 2017-02-14 | 2021-08-24 | Microsoft Technology Licensing, Llc | Intelligent device user interactions |
| CN113360653A (en) * | 2021-06-11 | 2021-09-07 | 湖北中烟工业有限责任公司 | Method and system for acquiring text intention |
| CN113672720A (en) * | 2021-09-14 | 2021-11-19 | 国网天津市电力公司 | Power audit question and answer method based on knowledge graph and semantic similarity |
| CN113688243A (en) * | 2021-08-31 | 2021-11-23 | 中国平安人寿保险股份有限公司 | Method, device and equipment for marking entities in sentences and storage medium |
| CN114218392A (en) * | 2022-02-22 | 2022-03-22 | 浙商期货有限公司 | Futures question-answer oriented user intention identification method and system |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101630312A (en) * | 2009-08-19 | 2010-01-20 | 腾讯科技（深圳）有限公司 | Clustering method for question sentences in question-and-answer platform and system thereof |
| CN101777042A (en) * | 2010-01-21 | 2010-07-14 | 西南科技大学 | Neural network and tag library-based statement similarity algorithm |
| CN101887443A (en) * | 2009-05-13 | 2010-11-17 | 华为技术有限公司 | Method and device for classifying texts |
| CN102289522A (en) * | 2011-09-19 | 2011-12-21 | 北京金和软件股份有限公司 | Method of intelligently classifying texts |
| CN102662976A (en) * | 2012-03-12 | 2012-09-12 | 浙江工业大学 | Text feature weighting method based on supervision |
| CN103246686A (en) * | 2012-02-14 | 2013-08-14 | 阿里巴巴集团控股有限公司 | Method and device for text classification, and method and device for characteristic processing of text classification |
| CN103425777A (en) * | 2013-08-15 | 2013-12-04 | 北京大学 | Intelligent short message classification and searching method based on improved Bayesian classification |
| US8744837B2 (en) * | 2010-10-25 | 2014-06-03 | Electronics And Telecommunications Research Institute | Question type and domain identifying apparatus and method |
| CN103902570A (en) * | 2012-12-27 | 2014-07-02 | 腾讯科技（深圳）有限公司 | Text classification feature extraction method, classification method and device |
| CN104050256A (en) * | 2014-06-13 | 2014-09-17 | 西安蒜泥电子科技有限责任公司 | Initiative study-based questioning and answering method and questioning and answering system adopting initiative study-based questioning and answering method |
| CN104063387A (en) * | 2013-03-19 | 2014-09-24 | 三星电子（中国）研发中心 | Device and method abstracting keywords in text |
| CN104281653A (en) * | 2014-09-16 | 2015-01-14 | 南京弘数信息科技有限公司 | Viewpoint mining method for ten million microblog texts |
| CN104361010A (en) * | 2014-10-11 | 2015-02-18 | 北京中搜网络技术股份有限公司 | Automatic classification method for correcting news classification |
-
2015
- 2015-12-02 CN CN201510874529.0A patent/CN105389307A/en active Pending
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101887443A (en) * | 2009-05-13 | 2010-11-17 | 华为技术有限公司 | Method and device for classifying texts |
| CN101630312A (en) * | 2009-08-19 | 2010-01-20 | 腾讯科技（深圳）有限公司 | Clustering method for question sentences in question-and-answer platform and system thereof |
| CN101777042A (en) * | 2010-01-21 | 2010-07-14 | 西南科技大学 | Neural network and tag library-based statement similarity algorithm |
| US8744837B2 (en) * | 2010-10-25 | 2014-06-03 | Electronics And Telecommunications Research Institute | Question type and domain identifying apparatus and method |
| CN102289522A (en) * | 2011-09-19 | 2011-12-21 | 北京金和软件股份有限公司 | Method of intelligently classifying texts |
| CN103246686A (en) * | 2012-02-14 | 2013-08-14 | 阿里巴巴集团控股有限公司 | Method and device for text classification, and method and device for characteristic processing of text classification |
| CN102662976A (en) * | 2012-03-12 | 2012-09-12 | 浙江工业大学 | Text feature weighting method based on supervision |
| CN103902570A (en) * | 2012-12-27 | 2014-07-02 | 腾讯科技（深圳）有限公司 | Text classification feature extraction method, classification method and device |
| CN104063387A (en) * | 2013-03-19 | 2014-09-24 | 三星电子（中国）研发中心 | Device and method abstracting keywords in text |
| CN103425777A (en) * | 2013-08-15 | 2013-12-04 | 北京大学 | Intelligent short message classification and searching method based on improved Bayesian classification |
| CN104050256A (en) * | 2014-06-13 | 2014-09-17 | 西安蒜泥电子科技有限责任公司 | Initiative study-based questioning and answering method and questioning and answering system adopting initiative study-based questioning and answering method |
| CN104281653A (en) * | 2014-09-16 | 2015-01-14 | 南京弘数信息科技有限公司 | Viewpoint mining method for ten million microblog texts |
| CN104361010A (en) * | 2014-10-11 | 2015-02-18 | 北京中搜网络技术股份有限公司 | Automatic classification method for correcting news classification |
Non-Patent Citations (1)
| Title |
|---|
| 陈涛 等: "文本分类中的特征降维方法综述", 《情报学报》 * |
Cited By (70)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107291685B (en) * | 2016-04-13 | 2020-10-13 | 北京大学 | Semantic recognition method and semantic recognition system |
| CN107291685A (en) * | 2016-04-13 | 2017-10-24 | 北京大学 | Method for recognizing semantics and semantics recognition system |
| CN106056154A (en) * | 2016-05-27 | 2016-10-26 | 大连楼兰科技股份有限公司 | Fault code recognition and classification method |
| CN105975460A (en) * | 2016-05-30 | 2016-09-28 | 上海智臻智能网络科技股份有限公司 | Question information processing method and device |
| CN105955965A (en) * | 2016-06-21 | 2016-09-21 | 上海智臻智能网络科技股份有限公司 | Question information processing method and device |
| CN108121722A (en) * | 2016-11-28 | 2018-06-05 | 渡鸦科技（北京）有限责任公司 | The construction method and device of knowledge base |
| CN106649694A (en) * | 2016-12-19 | 2017-05-10 | 北京云知声信息技术有限公司 | Method and device for identifying user's intention in voice interaction |
| CN106649696A (en) * | 2016-12-19 | 2017-05-10 | 北京云知声信息技术有限公司 | Information classification method and device |
| CN106649696B (en) * | 2016-12-19 | 2020-05-26 | 北京云知声信息技术有限公司 | Information classification method and device |
| CN106649694B (en) * | 2016-12-19 | 2020-05-26 | 北京云知声信息技术有限公司 | Method and device for determining user intention in voice interaction |
| CN108228559A (en) * | 2016-12-22 | 2018-06-29 | 苏宁云商集团股份有限公司 | A kind of human-computer interaction realization method and system for customer service |
| US10957311B2 (en) | 2017-02-14 | 2021-03-23 | Microsoft Technology Licensing, Llc | Parsers for deriving user intents |
| US11004446B2 (en) | 2017-02-14 | 2021-05-11 | Microsoft Technology Licensing, Llc | Alias resolving intelligent assistant computing device |
| CN110313154A (en) * | 2017-02-14 | 2019-10-08 | 微软技术许可有限责任公司 | Intelligent assistant with the information discrimination based on intention |
| US10984782B2 (en) | 2017-02-14 | 2021-04-20 | Microsoft Technology Licensing, Llc | Intelligent digital assistant system |
| CN110313154B (en) * | 2017-02-14 | 2021-12-07 | 微软技术许可有限责任公司 | Intelligent assistant with intent-based information resolution |
| US11010601B2 (en) | 2017-02-14 | 2021-05-18 | Microsoft Technology Licensing, Llc | Intelligent assistant device communicating non-verbal cues |
| US11017765B2 (en) | 2017-02-14 | 2021-05-25 | Microsoft Technology Licensing, Llc | Intelligent assistant with intent-based information resolution |
| US11194998B2 (en) | 2017-02-14 | 2021-12-07 | Microsoft Technology Licensing, Llc | Multi-user intelligent assistance |
| US11126825B2 (en) | 2017-02-14 | 2021-09-21 | Microsoft Technology Licensing, Llc | Natural language interaction for smart assistant |
| US11100384B2 (en) | 2017-02-14 | 2021-08-24 | Microsoft Technology Licensing, Llc | Intelligent device user interactions |
| CN108446286A (en) * | 2017-02-16 | 2018-08-24 | 阿里巴巴集团控股有限公司 | A kind of generation method, device and the server of the answer of natural language question sentence |
| CN107122404A (en) * | 2017-03-22 | 2017-09-01 | 北京晓数聚传媒科技有限公司 | A kind of user view data extracting method and device |
| CN107153672A (en) * | 2017-03-22 | 2017-09-12 | 中国科学院自动化研究所 | User mutual intension recognizing method and system based on Speech Act Theory |
| CN107193865A (en) * | 2017-04-06 | 2017-09-22 | 上海奔影网络科技有限公司 | Natural language is intended to understanding method and device in man-machine interaction |
| CN107193865B (en) * | 2017-04-06 | 2020-03-10 | 上海奔影网络科技有限公司 | Natural language intention understanding method and device in man-machine interaction |
| CN107329948A (en) * | 2017-05-23 | 2017-11-07 | 努比亚技术有限公司 | Sentence describes time of origin presumption method, equipment and the storage medium of event |
| CN107329948B (en) * | 2017-05-23 | 2020-07-28 | 南京风数智能科技有限公司 | Method, apparatus and storage medium for estimating occurrence time of event described in sentence |
| CN109388705B (en) * | 2017-08-07 | 2020-05-19 | 芋头科技（杭州）有限公司 | Text intention classification method |
| CN109388705A (en) * | 2017-08-07 | 2019-02-26 | 芋头科技（杭州）有限公司 | A kind of text intent classifier method |
| CN109815308B (en) * | 2017-10-31 | 2021-01-01 | 北京小度信息科技有限公司 | Method and device for determining intention recognition model and method and device for searching intention recognition |
| CN109815308A (en) * | 2017-10-31 | 2019-05-28 | 北京小度信息科技有限公司 | The determination of intention assessment model and retrieval intension recognizing method, device |
| CN107943791A (en) * | 2017-11-24 | 2018-04-20 | 北京奇虎科技有限公司 | A kind of recognition methods of refuse messages, device and mobile terminal |
| CN109933779A (en) * | 2017-12-18 | 2019-06-25 | 苏宁云商集团股份有限公司 | User's intension recognizing method and system |
| CN108153737A (en) * | 2017-12-30 | 2018-06-12 | 北京中关村科金技术有限公司 | A kind of method of semantic classification, system and dialog process system |
| CN108268617A (en) * | 2018-01-05 | 2018-07-10 | 阿里巴巴集团控股有限公司 | User view determines method and device |
| CN108334495A (en) * | 2018-01-30 | 2018-07-27 | 国家计算机网络与信息安全管理中心 | Short text similarity calculating method and system |
| CN108334496B (en) * | 2018-01-30 | 2020-06-12 | 中国科学院自动化研究所 | Man-machine conversation understanding method and system for specific field and related equipment |
| CN108334496A (en) * | 2018-01-30 | 2018-07-27 | 中国科学院自动化研究所 | Human-computer dialogue understanding method and system and relevant device for specific area |
| CN110619050B (en) * | 2018-06-20 | 2023-05-09 | 华为技术有限公司 | Intention recognition method and device |
| CN110619050A (en) * | 2018-06-20 | 2019-12-27 | 华为技术有限公司 | Intention recognition method and equipment |
| CN108920579A (en) * | 2018-06-22 | 2018-11-30 | 龙马智芯（珠海横琴）科技有限公司 | Classification method, device, terminal device and the storage medium of sentence |
| CN109033075B (en) * | 2018-06-29 | 2022-04-15 | 北京百度网讯科技有限公司 | Intention matching method and device, storage medium and terminal equipment |
| CN109033075A (en) * | 2018-06-29 | 2018-12-18 | 北京百度网讯科技有限公司 | It is intended to matched method, apparatus, storage medium and terminal device |
| CN110659175A (en) * | 2018-06-30 | 2020-01-07 | 中兴通讯股份有限公司 | Log trunk extraction method, log trunk classification method, log trunk extraction equipment and log trunk storage medium |
| CN109241261A (en) * | 2018-08-30 | 2019-01-18 | 武汉斗鱼网络科技有限公司 | User's intension recognizing method, device, mobile terminal and storage medium |
| CN109388695A (en) * | 2018-09-27 | 2019-02-26 | 深圳前海微众银行股份有限公司 | User's intension recognizing method, equipment and computer readable storage medium |
| CN109522393A (en) * | 2018-10-11 | 2019-03-26 | 平安科技（深圳）有限公司 | Intelligent answer method, apparatus, computer equipment and storage medium |
| CN109857841A (en) * | 2018-12-05 | 2019-06-07 | 厦门快商通信息技术有限公司 | A kind of FAQ question sentence Text similarity computing method and system |
| CN109815492A (en) * | 2019-01-04 | 2019-05-28 | 平安科技（深圳）有限公司 | A kind of intension recognizing method based on identification model, identification equipment and medium |
| WO2020140372A1 (en) * | 2019-01-04 | 2020-07-09 | 平安科技（深圳）有限公司 | Recognition model-based intention recognition method, recognition device, and medium |
| CN109766414A (en) * | 2019-01-18 | 2019-05-17 | 广东小天才科技有限公司 | Intention identification method and system |
| CN109949637A (en) * | 2019-03-13 | 2019-06-28 | 广东小天才科技有限公司 | Automatic answering method and device for objective questions |
| CN110232114A (en) * | 2019-05-06 | 2019-09-13 | 平安科技（深圳）有限公司 | Sentence intension recognizing method, device and computer readable storage medium |
| WO2020224213A1 (en) * | 2019-05-06 | 2020-11-12 | 平安科技（深圳）有限公司 | Sentence intent identification method, device, and computer readable storage medium |
| CN110457466A (en) * | 2019-06-28 | 2019-11-15 | 谭浩 | Generate method, computer readable storage medium and the terminal device of interview report |
| CN110909142B (en) * | 2019-11-20 | 2023-03-31 | 腾讯科技（深圳）有限公司 | Question and sentence processing method and device of question-answer model, electronic equipment and storage medium |
| CN110909142A (en) * | 2019-11-20 | 2020-03-24 | 腾讯科技（深圳）有限公司 | Question and sentence processing method and device of question-answer model, electronic equipment and storage medium |
| CN111507085A (en) * | 2019-11-25 | 2020-08-07 | 江苏艾佳家居用品有限公司 | Sentence pattern recognition method |
| CN111507085B (en) * | 2019-11-25 | 2023-07-07 | 江苏艾佳家居用品有限公司 | Sentence pattern recognition method |
| CN111128156A (en) * | 2019-12-10 | 2020-05-08 | 上海雷盎云智能技术有限公司 | Intelligent household equipment voice control method and device based on model training |
| CN112989839A (en) * | 2019-12-18 | 2021-06-18 | 中国科学院声学研究所 | Keyword feature-based intent recognition method and system embedded in language model |
| CN111159403B (en) * | 2019-12-27 | 2022-07-29 | 广东高乐教育科技有限公司 | Intelligent classroom perception method and system |
| CN111159403A (en) * | 2019-12-27 | 2020-05-15 | 广东高乐教育科技有限公司 | Intelligent classroom perception method and system |
| CN113360653A (en) * | 2021-06-11 | 2021-09-07 | 湖北中烟工业有限责任公司 | Method and system for acquiring text intention |
| CN113688243A (en) * | 2021-08-31 | 2021-11-23 | 中国平安人寿保险股份有限公司 | Method, device and equipment for marking entities in sentences and storage medium |
| CN113688243B (en) * | 2021-08-31 | 2024-02-13 | 中国平安人寿保险股份有限公司 | Method, device, equipment and storage medium for labeling entities in sentences |
| CN113672720A (en) * | 2021-09-14 | 2021-11-19 | 国网天津市电力公司 | Power audit question and answer method based on knowledge graph and semantic similarity |
| CN114218392A (en) * | 2022-02-22 | 2022-03-22 | 浙商期货有限公司 | Futures question-answer oriented user intention identification method and system |
| CN114218392B (en) * | 2022-02-22 | 2022-06-03 | 浙商期货有限公司 | Futures question-answer oriented user intention identification method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105389307A (en) | Statement intention category identification method and apparatus | |
| CN109472033B (en) | Method and system for extracting entity relationship in text, storage medium and electronic equipment | |
| CN108304468B (en) | Text classification method and text classification device | |
| US20190287142A1 (en) | Method, apparatus for evaluating review, device and storage medium | |
| US11468342B2 (en) | Systems and methods for generating and using knowledge graphs | |
| CN111159363A (en) | Knowledge base-based question answer determination method and device | |
| CN104834651A (en) | Method and apparatus for providing answers to frequently asked questions | |
| CN103377239A (en) | Method and device for calculating inter-textual similarity | |
| CN105512104A (en) | Dictionary dimension reducing method and device and information classifying method and device | |
| US20200364216A1 (en) | Method, apparatus and storage medium for updating model parameter | |
| CN108733675B (en) | Emotion evaluation method and device based on large amount of sample data | |
| CN105843796A (en) | Microblog emotional tendency analysis method and device | |
| CN107341143A (en) | A kind of sentence continuity determination methods and device and electronic equipment | |
| CN111782793A (en) | Intelligent customer service processing method, system and equipment | |
| CN114722176A (en) | Intelligent question answering method, device, medium and electronic equipment | |
| WO2020074022A1 (en) | Synonym search method and device | |
| CN114840685A (en) | Emergency plan knowledge graph construction method | |
| CN104572618A (en) | Question-answering system semantic-based similarity analyzing method, system and application | |
| CN113761104A (en) | Method and device for detecting entity relationship in knowledge graph and electronic equipment | |
| CN109299007A (en) | A kind of defect repair person's auto recommending method | |
| CN107291686B (en) | Method and system for identifying emotion identification | |
| CN116561298A (en) | Title generation method, device, equipment and storage medium based on artificial intelligence | |
| CN111209394A (en) | Text classification processing method and device | |
| CN110705258A (en) | Text entity identification method and device | |
| Chuang et al. | Resume parser: Semi-structured chinese document analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160309 |