CN103902652A - Automatic question-answering system - Google Patents
Automatic question-answering system Download PDFInfo
- Publication number
- CN103902652A CN103902652A CN201410068844.XA CN201410068844A CN103902652A CN 103902652 A CN103902652 A CN 103902652A CN 201410068844 A CN201410068844 A CN 201410068844A CN 103902652 A CN103902652 A CN 103902652A
- Authority
- CN
- China
- Prior art keywords
- answer
- question
- keyword
- answering system
- user
- 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
- 238000004458 analytical method Methods 0.000 claims abstract description 29
- 238000000605 extraction Methods 0.000 claims abstract description 23
- 239000000284 extract Substances 0.000 claims abstract description 21
- 230000003993 interaction Effects 0.000 claims abstract description 20
- 238000004422 calculation algorithm Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 11
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000000034 method Methods 0.000 description 39
- 230000000694 effects Effects 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 9
- 230000008569 process Effects 0.000 description 9
- 230000008878 coupling Effects 0.000 description 8
- 238000010168 coupling process Methods 0.000 description 8
- 238000005859 coupling reaction Methods 0.000 description 8
- 238000010276 construction Methods 0.000 description 4
- 238000003058 natural language processing Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000007812 deficiency Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000005192 partition Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 230000002123 temporal effect Effects 0.000 description 2
- 238000013519 translation Methods 0.000 description 2
- VYZAMTAEIAYCRO-UHFFFAOYSA-N Chromium Chemical compound [Cr] VYZAMTAEIAYCRO-UHFFFAOYSA-N 0.000 description 1
- 244000097202 Rathbunia alamosensis Species 0.000 description 1
- 235000009776 Rathbunia alamosensis Nutrition 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000011109 contamination Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000005309 stochastic process Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
- G06F40/35—Discourse or dialogue representation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- General Physics & Mathematics (AREA)
- Human Computer Interaction (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention discloses an automatic question-answering system. According to the automatic question-answering system, a user interaction unit receives a question input by a user, a question analysis unit extracts keywords from the question input by the user and expands the keywords, then an information retrieval unit searches a frequently asked question bank for answers to the question according to the expanded keywords and returns the answers to related documents, and finally, according to answer extraction rules corresponding to the question type, an answer extraction unit extracts an answer according with the rules from the related documents returned by the information retrieval unit, and sends the extracted answer to the user interaction unit to be fed back to the user. Compared with an automatic question-answering system in the prior art, the automatic question-answering system can automatically expand the keywords and divide the types of the input questions, different types of questions correspond to different types, further answer search is carried out in the types, and the accuracy and diversity of answers are improved.
Description
Technical field
The present invention relates to computer software technical field, particularly a kind of automatically request-answering system.
Background technology
Since the nineties in 20th century, Internet has worldwide obtained swift and violent development, and internet information is more and more, for people provide abundant information resources.On the other hand, online information is more and more, greatly promote the development of natural language processing technique, also natural language processing technique has been had higher requirement: people wish to obtain quickly and accurately in rambling network world the information of oneself wanting simultaneously.Although have on the internet a lot of search engines can help people to search for the information of oneself wanting existing, current search engine also has a lot of shortcomings, can not meet people easily and fast, the needs of obtaining information exactly.Show following three aspects: the one, correlation information is too many.The related web page that traditional search engine returns is too many, and user is difficult to navigate to rapidly and accurately required information.For example, user inputs several key words on Google, and it likely returns to thousands of webpages, and a lot of waste time in these webpages, is searched own needed information by user.The 2nd, with the incompatible expression Search Requirement of logical groups of keyword, because people's Search Requirement is very complicated and special often, to express with the simple combination of several keywords, user does not express clearly the retrieval intention of oneself like this, and search engine also just has no idea to find out to have made customer satisfaction system answer naturally.The 3rd, take keyword as basic index, matching algorithm, although this algorithm is simple, rests on after all the top layer of language, and do not touch semanteme, therefore retrieval effectiveness is difficult to further improve.
Automatic question answering (QA, Automatic Question Answering) technology is accompanied by natural language just semantic processes technology in order to meet people's this hope grows up.People can put question to automatically request-answering system with common question sentence, automatically request-answering system will be searched for corresponding answer from knowledge base or internet, then answer is in brief directly returned to user, rather than what as search engine, return to user is the webpage that a pile is relevant.User just can obtain the information of oneself wanting easily by automatically request-answering system like this.The technology such as the representation of knowledge, information retrieval, natural language processing have been used in automatic question answering technological synthesis.Automatically request-answering system can make user input problem with natural language, rather than crucial contamination.And return to user be succinctly, answer accurately, rather than some relevant webpages.So question answering system can better meet user's Search Requirement, can find out quickly the required answer of user.Can say, question answering system is exactly search engine of new generation.For question answering system, user does not need the PROBLEM DECOMPOSITION of oneself to become key word, and user can directly give question answering system whole problem.Question answering system, in conjunction with natural language processing technique, by problem is understood, can directly be submitted to the answer that user wants.Question answering system, just as an encyclopedic expert, can be answered any problem rapidly and accurately.Such as, user submit to a problem " what the abbreviation in Shanghai is? " question answering system will directly provide answer " abbreviation in Shanghai is Shanghai ".Can find out, question answering system is more convenient, fast, efficient than traditional search engine.
The automatically request-answering system of domestic not yet comparative maturity at present." a kind of Chinese natural language answering method based on question and answer storehouse " that prior art proposes, set up FAQ storehouse by internet professional website, then user is inquired about to participle analysis and obtain close inquiry question sentence, with question sentence retrieval question sentence, the mode of coupling answer, the foundation and the question sentence similarity that mainly comprise question and answer storehouse are calculated two aspects." a kind of automatically request-answering system and method " that Tencent Technology (Shenzhen) Co., Ltd. proposes provides the normalized unit to keyword in question answering system, make keyword in user's read statement can be converted into keyword general in inferenctial knowledge storehouse by normalized, thereby reduce the workload of building inferenctial knowledge storehouse." a kind of Intelligent Chinese-character question answering system based on concept " that the Central China University of Science and Technology proposes can be processed rear keyword string to the question sentence of user's input and carry out synonym expansion, better understand question sentence, retrieve, improve the recall ratio of question answering system, from morphology, word order, the long three aspects: of word has provided a kind of Chinese sentence similarity computing method based on concept, has improved precision ratio." FAQ Chinese request-answering system implementing method in tourism field " that Kunming University of Science and Technology proposes provides a kind of implementation method of tour field FAQ Chinese Question Answering System, comprises that FAQ collects and tissue, tour field construction of knowledge base, user's inquiry, the steps such as answer extraction.This realization, by ontological thought, has built tour field knowledge base-Domain-hownet, utilizes KDML language definition and has described tour field term and relation, and having proposed a kind of tourism question sentence similarity calculating method." a kind of automatic question-answering method and system " that Shenzhen Graduate School of Peking University proposes analyzed question sentence, the word of ask a little/condition point of employing model of cognition after to participle indicates, and utilizes the information resource database of asking point, condition point query SQL structure of the question sentence identifying to obtain result." a kind of server end of the method, device and the knowledge Q-A system that form enquirement " that Baidu In Line Network Technology Co Ltd (Beojing) proposes proposed to utilize puts question to template to obtain the mode that user inputs key message and puts question to." a kind of Questions &. Answers on Multimedia system and method " that Huawei Tech Co., Ltd proposes resolves according to user to user input problem, obtain characteristic information and semantic classes, in default multimedia database, search answer corresponding to problem that under this classification, similarity is the highest.
In above-mentioned technology, there is following problem:
The first, the question answering system based on common problem storehouse (FAQ), the scale of problem base and scope affect the accuracy of answer, so constructing the more comprehensive common problem of a ratio storehouse is the matter of utmost importance that such problem system need to solve, and within question answering system based on problem base is generally used for certain professional domain, its extendability is poor.In addition, user inputs question sentence and problem base is the core place of system with the similarity calculating between sentence, and the accuracy of its computing method and high efficiency are related to accuracy and the efficiency of whole system.
The second, the automatically request-answering system based on Internet, the information redundancy retrieving is excessive, may be subordinate to multiple subject informations, and the extraction process meeting more complicated of answer and the accuracy rate of answer can not be guaranteed.
The 3rd, keyword coupling and semantic extension problem in retrieval, but because expression way in Chinese is flexible, the position of appearance with identical semantic its keyword of sentence is also indefinite, keyword Match in sequence often can not meet retrieval requirement, in Chinese, exist a large amount of synonyms, in problem and answer, diverse keyword may contain identical semanteme, if not carrying out semantic extension also can cause and retrieve unsuccessfully, but having improved the recall rate of retrieval, semantic extension but may reduce the accuracy rate of retrieval, if what energy was intelligent marks off asked questions type, can improve true accurate rate.
Four, said method is with the problem in input problem retrieval FAQ storehouse mostly, return to the answer of the most similar problem in FAQ storehouse, for some matters of opening, within its answer is limited to certain field often, although can obtain the answer of one or more general character, reference is carried out in the answer that often cannot offer multiplicity of subscriber.
Summary of the invention
In view of the deficiencies in the prior art, the object of the invention is to provide one automatically input problem to be carried out to type division, searches for the automatically request-answering system of answer in the type.
Technical scheme of the present invention is as follows:
A kind of automatically request-answering system, comprising:
User interaction unit, for receiving the problem of user's input and problem answers being fed back to described user;
Question analysis unit, for extracting the keyword of problem of user input, and expands described keyword, and according to the Question Classification standard setting in advance, problem is carried out to type and divide the type that obtains described problem;
Frequently asked question storehouse, for storing problem and the answer of user Chang Wen;
Information retrieval unit, for according to the keyword after the expansion of described question analysis unit in the search problem answer of described frequently asked question storehouse, and return to relevant document or answer;
Answer extracting unit, extracts for the relevant documentation returning from described information retrieval unit according to the answer decimation rule corresponding with the type of described problem the answer that meets described rule, and the answer of extraction is sent to described user interaction unit.
Beneficial effect:
In the disclosed automatically request-answering system of the embodiment of the present invention, user interaction unit receives the problem of user's input, question analysis unit is carried out extracting keywords and keyword is expanded the problem of user's input, then by information retrieval unit according to expansion after keyword search problem answer in frequently asked question storehouse, find direct answer to be sent to described user interaction unit to user feedback, otherwise return to relevant document, the relevant documentation finally being returned from described information retrieval unit according to the answer decimation rule corresponding with described problem types by answer extracting unit, extract the answer that meets described rule, the answer of extraction is sent to described user interaction unit to user feedback.Compared with automatically request-answering system of the prior art, the automatically request-answering system that the embodiment of the present invention provides can automatically be expanded keyword and input problem be carried out to type division, dissimilar type corresponding to problem, and then in the type, search for answer, improve accuracy rate and the diversity of answer.Input problem carries out problem domain and type is divided, and divides by field, under this territory, searches for, dwindle hunting zone, the matched rule that provides answer to extract is provided by type, is returned to answer corresponding to the type according to rule, improved accuracy rate and the diversity of answer.
Accompanying drawing explanation
Below in conjunction with drawings and Examples, the invention will be further described, in accompanying drawing:
The structural representation of the automatically request-answering system that Fig. 1 provides for the embodiment of the present invention.
Embodiment
For making object of the present invention, technical scheme and effect clearer, clear and definite, below the present invention is described in more detail.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.
Consult Fig. 1, the structural representation of its automatically request-answering system providing for the embodiment of the present invention.As shown in Figure 1, described automatically request-answering system comprises:
User interaction unit 10, for receiving the problem of user's input and problem answers being fed back to described user;
Question analysis unit 20, for extracting the keyword of problem of user input, and expands described keyword, and according to the Question Classification standard setting in advance, problem is carried out to type and divide the type that obtains described problem;
Frequently asked question storehouse 30, for storing problem and the answer of user Chang Wen;
Information retrieval unit 40, for according to the keyword after 20 expansions of described question analysis unit in the search problem answer of described frequently asked question storehouse, and return to relevant document or answer;
Answer extracting unit 50, extracts for the relevant documentation returning from described information retrieval unit 40 according to the answer decimation rule corresponding with the type of described problem the answer that meets described rule, and the answer of extraction is sent to described user interaction unit 10.
In the disclosed automatically request-answering system of the embodiment of the present invention, user interaction unit 10 receives the problem of user's input, question analysis unit 20 is carried out extracting keywords and keyword is expanded the problem of user's input, then by information retrieval unit 40 according to expansion after keyword search problem answer in frequently asked question storehouse 30, and return to relevant document, the relevant documentation finally being returned from described information retrieval unit 40 according to the answer decimation rule corresponding with described problem types by answer extracting unit 50, extract the answer that meets described rule, the answer of extraction is sent to described user interaction unit 10 to user feedback.Compared with automatically request-answering system of the prior art, the automatically request-answering system that the embodiment of the present invention provides can automatically be expanded keyword and input problem be carried out to type division, dissimilar type corresponding to problem, and then in the type, search for answer, improve accuracy rate and the diversity of answer.
Understand in the automatically request-answering system providing in the embodiment of the present invention for more detailed, further introduce for the each functional module in automatically request-answering system below.
In embodiments of the present invention, user interaction unit 10 is user input query problem and the module of browsing problem answers, conventionally use browser, as conventional Internet Explorer, Firefox, Chrome browsers etc., the user's inquiry problem receiving is delivered to question analysis unit processing, or the problem answers obtaining from common problem storehouse feed back to user.
The automatically request-answering system that the embodiment of the present invention provides also comprises ontology knowledge storehouse 60, and described ontology knowledge storehouse provides shared vocabulary for described question analysis unit 20.Body is introduced automatically request-answering system by the embodiment of the present invention makes system carry out semantic analysis to user's query word, more can fully understand user's query intention, thereby effectively improve precision ratio and recall ratio.Ontology describing the concept in field and the semantic relation between them, can help machine domain knowledge to be had on the basis of profound understanding and do technical operation.Concrete, ontology knowledge storehouse refers to a kind of formal, for the clear and definite and detailed explanation of shared ideas system.Ontology knowledge storehouse provides a kind of vocabulary, namely those object types that exist or concept and attribute and mutual relationship among specific area shared; Ontology knowledge storehouse is a kind of terminology of specific type, has structurized feature, and is more suitable for using among computer system; In fact ontology knowledge storehouse is exactly to certain cover concept among specific area and the formalization expression (formal representation) of relation each other thereof.
Further, the system that the embodiment of the present invention provides has utilized the body of semantization as knowledge base, ontology describing the concept in field and the semantic relation between them, can help machine domain knowledge to be had on the basis of profound understanding and do technical operation.
Native system adopts a kind of full automatic body configuration method.The method does not need manual intervention, keyword that generally speaking only need to be a small amount of to each concept definition.Can automatically move subsequently, from network, crawl information, constructs body.The method can be divided into 4 steps to the training process of each Ontological concept:
Step (1) is used search engine inquiry relevant documentation;
Step (2) is according to the document in step (1), by LDA model generation candidate word;
Step (3) uses semantic range formula to give a mark to candidate word, as NGD, and WebJaccard etc.;
Step (4) joins the word after marking in the middle of the example of Ontological concept;
The first step of construction algorithm is to search relevant document for the each concept in body.For example, we can inquire about nearest hot news as relevant documentation from Baidu's news.
The second step of construction algorithm is to choose candidate word the document drawing from the first step.These candidate word should be words important in this document sets.Here we carry out keyword extraction with LDA (Latent Dirichlet Allocation) model.LDA is a kind of probability generation model of document sets.His basic thought is: each document is to be made up of a series of implicit themes of following certain distribution, and the word that may occur in each theme also has it specifically to distribute.After the weight of all words is all calculated, we sort to these words, and n(n=400 before selecting) individual word is as the candidate word of corresponding concept.
For the candidate word of choosing is marked more accurately, we use the formula that returns to number based on search engine inquiry to give a mark, and the most frequently used formula is: Normalized Google Distance (NGD).NGD is a formula of the close relation degree between two words being evaluated with Google Search Results.In our model, a given word w and a concept c, calculate NGD value between them and just must choose a word and represent this concept.Therefore we increase an attribute tag in body, represent the defined terms of each concept.
After finally choosing marking again, front n the highest candidate word of score joins in the middle of the example of Ontological concept, pass through said process, obtained concept and the example, but the relation between concept and concept and between concept and attribute not yet add, this can realize by HowNet.HowNet be one take the bilingual commonsense knowledge base as representative, its elementary organization unit is concept.Concept is used adopted former definition.The relation that relation, concept and the adopted former relation of concept and concept and justice are former and justice is former has formed the netted knowledge hierarchy of HowNet.HowNet defined hyponymy, synonymy, antonymy, to adopted relation, attribute-host relation, parts-whole relation, material-finished product relation, 8 kinds of relations of event-role relation, thereby complete the automatic construction process of body.
It is necessary analytical work before question answering system is retrieved that problem is understood, and the effect of this process analysis procedure analysis has important impact to processing procedure below.Problem is understood part need to complete following a few some work: extract the keyword of problem, according to factors such as the types of problem, keyword carried out to suitable expansion, classification under problem identificatioin, carries out pattern extraction by predefined question mode to problem and obtains problem types.If first the question answering system of Chinese will carry out participle and part-of-speech tagging etc. to problem.
Word is the least unit of information representation, and Chinese is different from western language, there is no separator (space) between the word of its sentence, therefore needs to carry out word and carries out cutting.The question analysis unit 20 that the embodiment of the present invention provides comprises: Chinese word processing module 21, and for the Chinese problem of user's input is carried out to word segmentation and part-of-speech tagging.
In Chinese terms cutting, there is cutting difference, if sentence " make user satisfied " can cutting be " making/user/satisfaction ", may be also " use/family/satisfaction " by cutting mistakenly, thereby need to utilize various Context Knowledges solution word segmentation differences.On the basis of cutting, utilize method rule-based and statistics (Markov chain) to carry out part-of-speech tagging.N-gram statistical analysis technique based on Markov chain stochastic process, is proved to be in part-of-speech tagging and can reaches higher precision.Here the participle program that the embodiment of the present invention is used is the Words partition system that do mechanical translation research department of School of Computer Science of Harbin Institute of Technology, it can disconnect each word in the Chinese language text of input, and after each word, indicates the part of speech of this word with a symbol.For example :/ng represents termini generales ,/nx represents Chinese surname, and/vg represents general verb etc.An example sentence through participle and part-of-speech tagging below:
Harbin/nd /p what/r place/ng? / wj
Further, the question analysis unit 20 that the embodiment of the present invention provides comprises keywords/concepts abstraction module 22, for extracting keyword according to the part of speech of the word after cutting.
In the problem that the embodiment of the present invention need to be putd question to user, extracted the useful key word of searching system below by keyword abstraction module 22.Be not that each word in problem can extract the keyword as searching system.Such as, interrogative with some are conventional ",, " etc. that word just should be filtered, for this reason, an inactive vocabulary of needs filter these words.
Keyword is mainly made up of noun, verb, adjective, limited adverbial word etc.Keyword can be divided into two kinds: the keyword of general keyword, " must contain ".The keyword of so-called " must contain " refers to these keywords and must in answer sentence, contain, and general keyword can not contained by Answer Sentence attached bag.Keyword is endowed different weights, and in the time of retrieval sentence, these weights are used for calculating the weight of sentence.Conventionally noun, the adverbial word with limited effect have higher weight.The keyword that " must contain " was made up of proper noun, limited adverbial word (as: maximum, the highest, the fastest etc.), time (as: 1997).The keyword principle that why will formulate " must contain " is because they have extremely strong limited effect to problem, is may be correct answer hardly if do not contain their sentence.For example: problem is " the highest Shi Nazuo mountain, mountain peak in the world? " and there is " Qiao Geli mountain is the second in the world peak " in the result of retrieval, this is not obviously the conceivable result of user, why occurs that the reason of this situation is just that very important keyword " the highest " is not contained by answer sentence.If add this restriction of keyword of " must contain ", this answer just can not be retrieved out so, therefore can greatly improve the accuracy of retrieval by the effect of these keywords.
The Chinese Question Answering System that the embodiment of the present invention provides also comprises the process that keyword is expanded, and in answer sentence, some word is not usually the keyword of original problem, but the expansion of the synonym of these words.For example: problem be " French Revolution which year occur? ", the sentence of answer is " 18 end of the centurys, the capitalist class great revolution of France's outburst." what in problem, use is " generation ", and in answer, used " outburst " this word.This has just caused keyword query failure, and therefore the embodiment of the present invention need to be carried out suitable expansion to keyword.Question analysis unit 20 in the automatically request-answering system that the embodiment of the present invention provides also comprises keyword expansion module 23, for the keyword extracting is carried out to synonym expansion.
Although keyword expansion has improved the recall rate of system, if expand inappropriate meeting and greatly reduce the accuracy rate of retrieval, therefore general question answering system is all very careful to the expansion of keyword.Here, the embodiment of the present invention is carried out keyword expansion from two aspects.First, the keyword using the synonym of all words as expansion; Secondly,, for the problem of some type, in corresponding answer, often there will be the word of certain common trait.For example,, for the problem in inquiry place, in answer, often there will be " " " being positioned at " keyword such as " be located in ".The embodiment of the present invention is also expanded these words as keyword.
For the expansion of the meaning of a word, that in English, commonly use is WordNet.WordNet provides relational network between word and near synonym set and near synonym set.By the hypernym pointer of word in WordNet, just can find out the far and near degree between word and word.For example, the hypernym pointer of " ex-husband " and " ex-wife " all points to " ex-spouse ", so these two words semantically can link together by " ex-spouse ".WordNet system is for English, for Chinese, and can be with knowing the semantic knowledge resource of net (HowNet) as system.Knowing that net is that a concept take Chinese and english representative is description object, is the commonsense knowledge base of substance with the pass of disclosing between concept and concept and between the attribute that concept was had.It is a netted organic knowledge system.
The Chinese Question Answering System that the embodiment of the present invention provides also comprises the process that problem types is divided, in embodiments of the present invention, each problem is often by specific type (being field), for example " when the People's Republic of China (PRC) sets up? " belong to historical field (history type), " Yao Ming's height is how many? " belong to sports field (sports genre); " how long the Changjiang river has? " belong to geographical field.In advance the frequently asked question in FAQ storehouse is carried out to the classification by field, then input problem is classified, then under such, retrieve, can effectively improve accuracy rate and the speed of retrieval.Current most automatically request-answering system is all to classify according to the classification providing in advance, but still there is the place of a lot of deficiencies in this classification, too many artificial factor, and classification is too thick, requirement that can not be completely realistic, for this problem, in embodiment of the present invention system, automatic classification method based on semantic will be used.Therefore, question analysis unit 20 in the automatically request-answering system that the embodiment of the present invention provides also comprises problem types division module 24, for divide the type of described problem according to the keyword after the Question Classification standard and the described expansion that set in advance, or for carry out automatically dividing the type of described problem according to the semanteme of the keyword after described expansion.
The Chinese Question Answering System that the embodiment of the present invention provides also comprises the process that question mode extracts, and the embodiment of the present invention is take Chinese Question Answering System as example here.General question answering system is the pattern extraction that carries out to problem according to query phrase all.Following table has been listed common problem types:
Table 1 FAQs type
Be directed to dissimilar problem and formulate corresponding answer decimation rule, to apply these rules and extract the answer of problem at answer extraction stage.Such as the problem for inquiry place, the embodiment of the present invention just can stipulate, in answer, must contain positional information.In system, by the automatic pattern extraction method using based on semantic, first collect a large amount of problems as corpus, then count the query phrase of frequent appearance by program.Such as passing through statistics, " what color " these words often appear in problem, and that embodiment of the present invention just can be used as a query phrase " what color ".Then every containing " what color " this phrase problem be all used as a class problem.
Common problem storehouse (FAQ) 20 refers to frequently asked question storehouse (Frequently-Asked Question).The effect in frequently asked question storehouse is that problem and answer that user is often asked save.Like this, for the problem of user's input, can first in FAQ storehouse, search for, looking at does not have identical problem.If had, just can directly answer corresponding to this problem in FAQ storehouse be returned to user.Like this, the problem of often asking for user, question answering system just can provide answer soon, and does not need through complicated below processing procedure, has so just greatly improved the efficiency of system.
In embodiments of the present invention, automatically request-answering system comprises that 30YuFAQ storehouse, FAQ storehouse updating block forms, frequently asked question storehouse updating block, for the answer from problem described in internet hunt in the time that described information retrieval unit does not search problem answers in described frequently asked question storehouse, and the answer searching is added to described frequently asked question storehouse.
The foundation in initial FAQ storehouse relies on the related resource on internet, uses web crawlers to crawl the question and answer information in FAQ and Ask-Answer Community on webpage, and preserves after structural data is processed.The webpage that contains FAQ can be searched by the search engine such as Google, Baidu, method be in these search engines take " inurl:faq " as querying condition, what search engine returned is the webpage that contains FAQ.Ask-Answer Community, existing internet comprises that Baidu is known, Sina likes to ask etc., these open communities allow domestic consumer to browse their problem and corresponding answer, the organizational form of question and answer on the page fixed, and correct option is had to clear and definite mark, is easy to extract.The updating block of FAQ refers to the problem that does not have user to ask in FAQ storehouse, the invention process regular meeting extracts by information retrieval and the answer of system the answer of finding coupling from Internet so, obtains this problem and the corresponding answer that after answer, just user can be asked and adds FAQ storehouse.
In embodiments of the present invention, the task of information retrieval unit 40 is exactly in document library, to search relevant document or answer with the key word extracting above.What information retrieval unit 40 was returned is some maximally related documents.Information retrieval unit 40 in question answering system also can directly be called existing searching system, such as Smart system, or also can call search engine on Internet such as Google.The input of information retrieval unit 40 is all generally the combination of key word, if English question answering system also needs key word to carry out root operation (Stemming).
Set up an information retrieval unit 40, need to set up index to document library.Could find rapidly like this document that comprises particular keywords.Before setting up index, be necessary language material to carry out pre-service, such as removing the document repeating, if English language material need to carry out root operation (Stemming), if Chinese data needs participle.
Key in information retrieval unit 40 is determining and document is sorted document weight.What the weight of document was commonly used is TF-IDF algorithm, and formula is as follows:
Wherein: KWi is i keyword comprising of the document in the weight in case study stage, and TFi is the frequency that this keyword occurs in this piece of document, and IDFi is the anti-frequency that this keyword occurs in document, and D refers to the distribution density of key word in document.Higher its TF is just larger for frequency that keyword occurs in the document, and keyword occurs that in more documents its IDF is just less, otherwise larger, it is more concentrated that keyword distributes in this piece of document, and D value is larger.TF*IDF value has reflected the significance level of this keyword from an aspect, conventionally in a document, often occur the word of (TF is large), and the word (IDF is large) in other documents seldom now, and the contained quantity of information of this word is just more, and this word is also just more important.If the distribution of keyword in document is more intensive in addition, this piece of document package is larger containing the possibility of associated answer, and the weight of this piece of document is just larger.After the complete weight of document calculations, just can sort to document according to weight, those documents of weight maximum are returned to answer extracting unit 50.
What general search engine returned is a pile document (webpage), is brief answer and question answering system need to be returned.So the answer extracting unit 50 in embodiment of the present invention automatically request-answering system comprises:
Relevance ranking module, carries out relevance ranking for the relevant documentation that information retrieval unit 40 is returned, and obtains the high document of correlativity; Document abstraction module 52, for extracting from the high document of described correlativity the one or more answers that meet described rule according to the answer decimation rule corresponding with described problem types; And answer sorting module 53, carry out cluster for multiple answers that described answer extracting unit 50 is extracted, the answer after cluster is sent to described user interaction unit.Wherein, the object of carrying out cluster that checks on one's answers in the embodiment of the present invention is in order to allow system can return to as much as possible diversified answer, thereby meets to greatest extent user's enquirement requirement.
Relevance ranking module first will, for example, to the document returning (webpage) according to its relevance ranking, be taken out the wherein higher document of correlativity, and the document abstraction module 52 of giving in answer extracting unit 50 refines one or more answers.The form of answer can be word, sentence, paragraph or digest.If the form of answer is word, sentence or paragraph, deal with also fairly simplely, if the form of answer is digest, so just need to use many Document Automatic Summarization Techniques.After being drawn into multiple candidate answers, the answer sorting module 53 in the embodiment of the present invention can be used the method for cluster, checks on one's answers and arranges, and feeds back to user interaction unit 10 according to the correlativity of answer and multifarious balance.
Different problems often has different answer forms and different answer abstracting methods.Therefore need every class problem to formulate an answer decimation rule.According to the type of problem, the form of answer can be word, sentence, paragraph or digest.In addition, for some problem types, answer must meet specific condition.Table 2 is and the extraction answer rule of table 1 correspondence.
Table 2 answer decimation rule
If using sentence as answer, deal with relatively simple, document high correlativity is divided into sentence by above-mentioned document abstraction module 52, calculate the weight of each sentence and sort and obtain candidate answers according to weight according to TF-IDF algorithm, according to described problem types, described candidate answers being sorted and obtains the answer of described problem.But, the problem of asking time place for those, its answer is just more brief, and does not need in short.Such as, for problem: " when the People's Republic of China (PRC) sets up? " the answer that the embodiment of the present invention is expected is a phrase, i.e. " on October 1st, 1949 ".But the embodiment of the present invention may retrieve such a word: " since on October 1st, 1949 the founding of People's Republic of China to end in 1994; China has set up the diplomatic relations with approximately 160 countries in the world, but also has developed economic and trade ties and culture contact with more countries and regions." can find out from this example; the desired answer of the embodiment of the present invention is the sub-fraction in the words; if the embodiment of the present invention can all be submitted to user as answer using this whole word, and obviously redundant information is too many, so phrase answer need to be extracted.
For the convenience of processing, what a lot of question answering systems was returned is that sentence is as answer.In this system, the detailed step of the extraction of answer is as follows:
(1) document being retrieved is divided into sentence
(2), according to certain algorithm, calculate the weight of each sentence
(3) sentence is sorted according to weight
(4) according to the type of problem, candidate answers is resequenced
The weight of calculating sentence in step (2) still adopts the algorithm of above-mentioned TF-IDF, after sorting, also needs the type according to problem to resequence to candidate answers according to weight.The special requirement that checks on one's answers of every class problem, so every class problem has own specific answer decimation rule.Rearrangement in step (4) is carried out according to these rules.For the problem of time correlation, in answer, just must contain temporal information.For containing numerical information in the relevant problem answers of quantity, otherwise just can not be correct option.
In addition, ask the problem in time place for those, can use very short statement answer, and for some problem, a brief phrase or in short clear hardly, such as for picture " what is it about 9.11 events? " this problem, has many relevant reports on the internet, and these reports may be to describe this event from different aspects.If all give user these relevant reports, user will spend a lot of time to read so.If these relevant reports can be made to a brief digest, as long as allow user see that digest just can know the cause and effect of whole event, will bring great convenience for user so.This just need to use many Document Automatic Summarization Techniques, and the document abstraction module 52 in the embodiment of the present invention extracts the content of common concern in the document that each correlativity is high as the answer of described problem.The relevant documentation that many Document Automatic Summarizations module can be retrieved information retrieval unit 40 is made digest and is returned to user.
The basic thought of many Document Automatic Summarizations is exactly to extract the common main contents of paying close attention in each document, eliminates redundant information identical between each document, then generates digest by certain algorithm.Many Document Automatic Summarizations can be found out by the cluster of sentence the theme of common concern, and what the sentence of getting together was often described is identical problem, and what can say a class representative is a theme.The sentence quantity of getting together is more, illustrates that the importance of this main body is larger, and then in important theme, selects the most representative sentence and form digest.
Understand in the automatically request-answering system providing in the embodiment of the present invention for more detailed, below for example to the each functional module in automatically request-answering system is further introduced.
User is by user interaction unit 10 input inquiry question sentences, as " Harbin somewhere? "
After question analysis unit 20 receives this problem, because word is the least unit of information representation, and Chinese is different from western language, there is no separator (space) between the word of its sentence, therefore needs to carry out word and carries out cutting.
First this question sentence is carried out to participle and semantic tagger, the participle program using is the Words partition system that do mechanical translation research department of School of Computer Science of Harbin Institute of Technology, it can disconnect each word in the Chinese language text of input, and after each word, indicates the part of speech of this word with a symbol.For example :/ng represents termini generales ,/nx represents Chinese surname, and/vg represents general verb etc.An example sentence through participle and part-of-speech tagging below:
Harbin/nd /p what/r place/ng? / wj
Keyword abstraction module 22 in question analysis unit 20 extracts keyword according to part-of-speech tagging, and keyword is mainly made up of noun, verb, adjective, limited adverbial word etc.Keyword can be divided into two kinds: the keyword of general keyword, " must contain ".The keyword of so-called " must contain " refers to these keywords and must in answer sentence, contain, and general keyword can not contained by Answer Sentence attached bag.Keyword is endowed different weights, and in the time of retrieval sentence, these weights are used for calculating the weight of sentence.Conventionally noun, the adverbial word with limited effect have higher weight.The keyword that " must contain " was made up of proper noun, limited adverbial word (as: maximum, the highest, the fastest etc.), time (as: 1997).The keyword principle that why will formulate " must contain " is because they have extremely strong limited effect to problem, is may be correct answer hardly if do not contain their sentence." Harbin " in above-mentioned question sentence, " ", " place ".
Keyword expansion module 23 in question analysis unit 20 is carried out synonym expansion to keyword, finds the synonym of some keyword.For Chinese, can be with knowing the semantic knowledge resource of net (HowNet) as system.Knowing that net is that a concept take Chinese and english representative is description object, is the commonsense knowledge base of substance with the pass of disclosing between concept and concept and between the attribute that concept was had.It is a netted organic knowledge system.In above-mentioned keyword " " and " being positioned at " belong to synonym, can do synonym expansion, and " being arranged in " " " occur manyly writtening language, express the meaning more clear.
Problem types in question analysis unit 20 is divided module 24, find out user and put question to the ken at place according to the keyword that extracts, in this example according to " Harbin ", " "/" being positioned at ", geographical knowledge field can be divided into, the scope of embodiment of the present invention search problem in FAQ storehouse can be dwindled by such division.
Question mode extracts according to the decimation rule making, as shown in table 1, in problem " where " problem types can be orientated as to inquiry place, to finally complete the decimation rule that obtains answer as table 2 according to this type, according to rule, in the answer that obtains extracting, must comprise location information.
After the processing of question analysis unit 20 complete dual problems, problem is submitted to frequently asked question storehouse (FAQ storehouse) 30, FAQ storehouse 30 can be retrieved by the mode of inverted index, also can pass through structural data, and SQL database mode is retrieved.Information retrieval unit 40 in the embodiment of the present invention is used invert indexed or SQL database retrieval FAQ storehouse 30, the question form of definition in the following example:
Like this to the problems referred to above, mate or keyword coupling can find corresponding problem by former problem, can be according to problem, the answer that obtains problem returns to user.
If former problem coupling or keyword coupling all cannot find corresponding Question ID, this problem is not in frequently asked question storehouse, need to go this problem of interconnected internet retrieval by search engine, obtain corresponding document, and the required answer of the embodiment of the present invention is just in candidate's document.
Answer extraction unit 50 carries out the operation of problem answers extraction to above-mentioned document.
First by the relevancy ranking unit in answer extraction unit 50, the document obtaining is carried out to relevancy ranking, then get the part in the forward document of correlativity rank wherein, as the document sets that proposes candidate answers.
Answer extraction unit 50 is according to the classification of question mode extracting unit to problem types before, obtain corresponding decimation rule, as shown in table 2, for example, to " when the People's Republic of China (PRC) sets up " this problem, according to table 2, the phrase that answer must contain temporal information is if " on October 1st, 1949 " or sentence are as " the founding of People's Republic of China was on October 1st, 1949 ", according to this rule, the phrase or the sentence that in document, meet above-mentioned condition are extracted, obtain candidate answers collection.
For another example for problem " what is it about 9.11 events? " for this problem, there are on the internet many relevant reports, these reports may be to describe this event from different aspects.If all give user these relevant reports, user will spend a lot of time to read so.If these relevant reports can be made to a brief digest, as long as allow user see that digest just can know the cause and effect of whole event, will bring great convenience for user so.This just need to use many Document Automatic Summarization Techniques.The relevant documentation that many Document Automatic Summarization Techniques can be retrieved information searching module is made digest as candidate answers.
Above-mentioned candidate answers is concentrated, and may comprise multiple possible answers, and the type that can obtain answer by cluster is divided, and for the unique problem of answer, by cluster, generally assembles out a class, then can return to user's answer.For some problem for example " when apple is set up? " answer is not unique, because " apple " can be IT industry " Apple (Apple Inc.) ", also can be " the apple group company " of Clothing Industry, can also be " apple record company (the Apple Record) " of music industry, assemble out at least 3 classes for such candidate answers, and each answer may be that user expects, so can not only return to one of them, can consider to feed back to the answer of multiplicity of subscriber.
The automatically request-answering system providing in the embodiment of the present invention. according to body, problem is classified, not merely domanial hierarchy just, and be concept hierarchy, for example basketball/NBA/NBA soccer star of star/the Lakers.
In addition, the system in the embodiment of the present invention is extracted the key comprising in problem, and for key, we also can do the normalized processing of synonym under the help of body.
The automatically request-answering system that the embodiment of the present invention provides, in the time of the ask a question pattern that extracts and corresponding answer regular, according to this pattern and rule learning and extract answer at next step, improves the accuracy rate that answer is extracted; For example, in order to improve the accuracy rate of coupling, our asking a little and condition point in can extraction problem, then does the pattern of the inquiry of being a problem, and extracts answer according to this pattern.
The question and answer answer that embodiment of the present invention automatically request-answering system obtains is document, but the embodiment of the present invention can be done further processing, it after extraction, can be word/phrase, sentence, even can be according to the digest extracting in multiple documents, the method that proposes to adopt multi-document auto-abstracting is mainly the time of reading answer in order to save user.
Automatically request-answering system of the prior art substantially all clearly do not ask a question answer variation consider, and automatically request-answering system in the embodiment of the present invention can to the answer of coupling according to correlativity and the multifarious weight that recalculates, thereby reach more desirable sequence, meet the demand of diversity and correlativity, feed back to user's answer.Here variation refers to just to be considered when calculating the sequence of answer, rather than check on one's answers carry out cluster this part.
Compared with prior art, the embodiment of the present invention is divided processing and pattern extraction by the problem in input problem and FAQ storehouse being carried out to type, the user's answer context obtaining is extensive, can comprise phrase, sentence in form, can also draw summary by multi-document summary algorithm, to candidate answers processing, improve accuracy rate and the diversity of answer by methods such as clusters.In addition, the embodiment of the present invention can make system carry out semantic analysis to user's query word by body being introduced to automatically request-answering system, more can fully understand user's query intention, thereby effectively improve precision ratio and recall ratio.
Should be understood that, application of the present invention is not limited to above-mentioned giving an example, and for those of ordinary skills, can be improved according to the above description or convert, and all these improvement and conversion all should belong to the protection domain of claims of the present invention.
Claims (10)
1. an automatically request-answering system, is characterized in that, comprising: user interaction unit, question analysis unit, frequently asked question storehouse, information retrieval unit and answer extracting unit,
Wherein, user interaction unit, for receiving the problem of user's input and problem answers being fed back to described user;
Question analysis unit, for extracting the keyword of problem of user input, and expands described keyword, and according to the Question Classification standard setting in advance, problem is carried out to type and divide the type that obtains described problem;
Frequently asked question storehouse, for storing problem and the answer of user Chang Wen;
Information retrieval unit, for according to the keyword after the expansion of described question analysis unit in the search problem answer of described frequently asked question storehouse, and return to relevant document or answer;
Answer extracting unit, extracts for the relevant documentation returning from described information retrieval unit according to the answer decimation rule corresponding with the type of described problem the answer that meets described rule, and the answer of extraction is sent to described user interaction unit.
2. automatically request-answering system according to claim 1, is characterized in that, described question analysis unit comprises:
Chinese word processing module, for carrying out word segmentation and part-of-speech tagging to the Chinese problem of user's input;
Keyword abstraction module, for extracting keyword according to the part of speech of the word after cutting;
Keyword expansion module, for carrying out synonym expansion to the keyword extracting;
Problem types is divided module, for divide the type of described problem according to the keyword after the Question Classification standard and the described expansion that set in advance, or for carry out automatically dividing the type of described problem according to the semanteme of the keyword after described expansion.
3. automatically request-answering system according to claim 2, is characterized in that, described answer extracting unit comprises:
Relevance ranking module, carries out relevance ranking for the relevant documentation that information retrieval unit is returned, and obtains the high document of correlativity;
Document abstraction module, for extracting from the high document of described correlativity the one or more answers that meet described rule according to the answer decimation rule corresponding with described problem types;
Answer sorting module, carries out cluster for multiple answers that described answer extracting unit is extracted, and the answer after cluster is sent to described user interaction unit.
4. automatically request-answering system according to claim 3, is characterized in that, if answer decimation rule corresponding to described problem types is sentence or paragraph,
Document high correlativity is divided into sentence by described document abstraction module, calculate the weight of each sentence and sort and obtain candidate answers according to weight according to TF-IDF algorithm, according to described problem types, described candidate answers being sorted and obtains the answer of described problem.
5. automatically request-answering system according to claim 3, is characterized in that, if answer decimation rule corresponding to described problem types is digest,
Described document abstraction module extracts the content of common concern in the document that each correlativity is high as the answer of described problem.
6. automatically request-answering system according to claim 2, is characterized in that, described information retrieval unit is by mode search problem answer in described frequently asked question storehouse of invert indexed or structural data.
7. according to the automatically request-answering system described in claim 1-6 any one, it is characterized in that, described system also comprises:
Frequently asked question storehouse updating block, for the answer from problem described in internet hunt in the time that described information retrieval unit does not search problem answers in described frequently asked question storehouse, and adds described frequently asked question storehouse by the answer searching.
8. according to the automatically request-answering system described in claim 1-6 any one, it is characterized in that, described user interaction unit is browser.
9. according to the automatically request-answering system described in claim 1-6 any one, it is characterized in that, the keyword that described question analysis unit extracts comprises noun, verb, adjective or adverbial word.
10. according to the automatically request-answering system described in claim 1-6 any one, it is characterized in that, described system also comprises ontology knowledge storehouse, and described ontology knowledge storehouse provides shared vocabulary for described question analysis unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410068844.XA CN103902652A (en) | 2014-02-27 | 2014-02-27 | Automatic question-answering system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410068844.XA CN103902652A (en) | 2014-02-27 | 2014-02-27 | Automatic question-answering system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103902652A true CN103902652A (en) | 2014-07-02 |
Family
ID=50993975
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410068844.XA Pending CN103902652A (en) | 2014-02-27 | 2014-02-27 | Automatic question-answering system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103902652A (en) |
Cited By (113)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102721A (en) * | 2014-07-18 | 2014-10-15 | 百度在线网络技术（北京）有限公司 | Method and device for recommending information |
| CN104461525A (en) * | 2014-11-27 | 2015-03-25 | 韩慧健 | Intelligent user-defined consulting platform generating system |
| CN104536991A (en) * | 2014-12-10 | 2015-04-22 | 乐娟 | Answer extraction method and device |
| CN104679815A (en) * | 2014-12-08 | 2015-06-03 | 北京云知声信息技术有限公司 | Method and system for screening question and answer pairs and updating question and answer database in real time |
| CN104778257A (en) * | 2015-04-20 | 2015-07-15 | 百度在线网络技术（北京）有限公司 | Method and device for searching applied problems |
| CN104820694A (en) * | 2015-04-28 | 2015-08-05 | 中国科学院自动化研究所 | Automatic Q&A method and system based on multi-knowledge base and integral linear programming ILP |
| CN104951433A (en) * | 2015-06-24 | 2015-09-30 | 北京京东尚科信息技术有限公司 | Method and system for intention recognition based on context |
| CN105095195A (en) * | 2015-07-03 | 2015-11-25 | 北京京东尚科信息技术有限公司 | Method and system for human-machine questioning and answering based on knowledge graph |
| CN105117387A (en) * | 2015-09-21 | 2015-12-02 | 上海智臻智能网络科技股份有限公司 | Intelligent robot interaction system |
| CN105117388A (en) * | 2015-09-21 | 2015-12-02 | 上海智臻智能网络科技股份有限公司 | Intelligent robot interaction system |
| CN105279274A (en) * | 2015-10-30 | 2016-01-27 | 北京京东尚科信息技术有限公司 | Answer combining and matching method and system based on natural synthetic answer system |
| CN105335400A (en) * | 2014-07-22 | 2016-02-17 | 阿里巴巴集团控股有限公司 | Method and apparatus for obtaining answer information for questioning intention of user |
| CN105354180A (en) * | 2015-08-26 | 2016-02-24 | 欧阳江 | Method and system for realizing open semantic interaction service |
| CN105512349A (en) * | 2016-02-23 | 2016-04-20 | 首都师范大学 | Question and answer method and question and answer device for adaptive learning of learners |
| CN105574133A (en) * | 2015-12-15 | 2016-05-11 | 苏州贝多环保技术有限公司 | Multi-mode intelligent question answering system and method |
| CN105630938A (en) * | 2015-12-23 | 2016-06-01 | 深圳市智客网络科技有限公司 | Intelligent question-answering system |
| CN105630917A (en) * | 2015-12-22 | 2016-06-01 | 成都小多科技有限公司 | Intelligent answering method and intelligent answering device |
| CN105630887A (en) * | 2015-12-18 | 2016-06-01 | 北京中科汇联科技股份有限公司 | Representation method for knowledge markup languages of Chinese question answering system and Chinese question answering system |
| CN105653619A (en) * | 2015-12-25 | 2016-06-08 | 上海智臻智能网络科技股份有限公司 | Update method and device of correct log library in intelligent question-answering system |
| CN105740310A (en) * | 2015-12-21 | 2016-07-06 | 哈尔滨工业大学 | Automatic answer summarizing method and system for question answering system |
| CN105824933A (en) * | 2016-03-18 | 2016-08-03 | 苏州大学 | Automatic question answering system based on main statement position and implementation method thereof |
| CN105893552A (en) * | 2016-03-31 | 2016-08-24 | 成都小多科技有限公司 | Method and device for processing data |
| CN105893465A (en) * | 2016-03-28 | 2016-08-24 | 北京京东尚科信息技术有限公司 | Automatic question answering method and device |
| CN105912527A (en) * | 2016-04-19 | 2016-08-31 | 北京高地信息技术有限公司 | Method, device and system outputting answer according to natural language |
| CN105912697A (en) * | 2016-04-25 | 2016-08-31 | 北京光年无限科技有限公司 | Optimization method and device of dialog system knowledge base |
| CN105955976A (en) * | 2016-04-15 | 2016-09-21 | 中国工商银行股份有限公司 | Automatic answering system and method |
| CN106095965A (en) * | 2016-06-17 | 2016-11-09 | 上海智臻智能网络科技股份有限公司 | A kind of data processing method and device |
| CN106202288A (en) * | 2016-06-30 | 2016-12-07 | 北京智能管家科技有限公司 | The optimization method of a kind of man-machine interactive system knowledge base and system |
| CN106547785A (en) * | 2015-09-22 | 2017-03-29 | 阿里巴巴集团控股有限公司 | Information getting method and system in knowledge base |
| CN106599297A (en) * | 2016-12-28 | 2017-04-26 | 北京百度网讯科技有限公司 | Method and device for searching question-type search terms on basis of deep questions and answers |
| CN106649394A (en) * | 2015-11-03 | 2017-05-10 | 中兴通讯股份有限公司 | Fusion knowledge base processing method and device and knowledge base management system |
| CN106878819A (en) * | 2017-01-20 | 2017-06-20 | 合网络技术（北京）有限公司 | The method, system and device of information exchange in a kind of network direct broadcasting |
| CN106934012A (en) * | 2017-03-10 | 2017-07-07 | 上海数眼科技发展有限公司 | A kind of question answering in natural language method and system of knowledge based collection of illustrative plates |
| CN106951470A (en) * | 2017-03-03 | 2017-07-14 | 中兴耀维科技江苏有限公司 | A kind of intelligent Answer System retrieved based on professional knowledge figure |
| CN106980624A (en) * | 2016-01-18 | 2017-07-25 | 阿里巴巴集团控股有限公司 | A kind for the treatment of method and apparatus of text data |
| CN106991161A (en) * | 2017-03-31 | 2017-07-28 | 北京字节跳动科技有限公司 | A kind of method for automatically generating open-ended question answer |
| CN107133210A (en) * | 2017-04-20 | 2017-09-05 | 中国科学院上海高等研究院 | Scheme document creation method and system |
| CN107145510A (en) * | 2017-03-31 | 2017-09-08 | 西安科技大学 | A kind of mathematical formulae searching method and device |
| CN107168599A (en) * | 2016-03-07 | 2017-09-15 | 阿里巴巴集团控股有限公司 | A kind of instant communicating method and device |
| CN107180037A (en) * | 2016-03-09 | 2017-09-19 | 北京京东尚科信息技术有限公司 | Man-machine interaction method and device |
| CN107229675A (en) * | 2017-04-28 | 2017-10-03 | 北京神州泰岳软件股份有限公司 | Question and answer base construction method, method of answering, the apparatus and system of list type knowledge |
| CN107301213A (en) * | 2017-06-09 | 2017-10-27 | 腾讯科技（深圳）有限公司 | Intelligent answer method and device |
| CN107330040A (en) * | 2017-06-27 | 2017-11-07 | 李博 | One kind study topic searching method and its system |
| CN107341157A (en) * | 2016-04-29 | 2017-11-10 | 阿里巴巴集团控股有限公司 | A kind of customer service dialogue clustering method and device |
| CN107408110A (en) * | 2015-03-06 | 2017-11-28 | 国立研究开发法人情报通信研究机构 | Meaning pairing expansion device, computer program therefor, and question-response system |
| CN107436916A (en) * | 2017-06-15 | 2017-12-05 | 百度在线网络技术（北京）有限公司 | The method and device of intelligent prompt answer |
| CN107453980A (en) * | 2017-07-26 | 2017-12-08 | 北京小米移动软件有限公司 | Problem response method and device in instant messaging |
| CN107528761A (en) * | 2016-06-21 | 2017-12-29 | 天脉聚源（北京）科技有限公司 | A kind of method and system of wechat public number automatic information reply |
| CN107729549A (en) * | 2017-10-31 | 2018-02-23 | 深圳追科技有限公司 | A kind of robot client service method and system comprising elements recognition |
| CN108073664A (en) * | 2016-11-11 | 2018-05-25 | 北京搜狗科技发展有限公司 | A kind of information processing method, device, equipment and client device |
| CN108170780A (en) * | 2017-12-26 | 2018-06-15 | 北京邦邦共赢网络科技有限公司 | A kind of the problem of self-service question and answer matching process and device |
| CN108170679A (en) * | 2017-12-28 | 2018-06-15 | 中国联合网络通信集团有限公司 | It can recognize that the semantic matching method and system of natural language description based on computer |
| WO2018121759A1 (en) * | 2016-12-31 | 2018-07-05 | 深圳市优必选科技有限公司 | Intelligent question and answer method and system |
| CN108416033A (en) * | 2018-03-12 | 2018-08-17 | 北京汽车研究总院有限公司 | A kind of data analysing method, device and terminal for automobile research and development |
| WO2018157805A1 (en) * | 2017-03-03 | 2018-09-07 | 腾讯科技（深圳）有限公司 | Automatic questioning and answering processing method and automatic questioning and answering system |
| CN108595696A (en) * | 2018-05-09 | 2018-09-28 | 长沙学院 | A kind of human-computer interaction intelligent answering method and system based on cloud platform |
| CN108595494A (en) * | 2018-03-15 | 2018-09-28 | 腾讯科技（深圳）有限公司 | The acquisition methods and device of reply message |
| CN108763529A (en) * | 2018-05-31 | 2018-11-06 | 苏州大学 | A kind of intelligent search method, device and computer readable storage medium |
| CN108804627A (en) * | 2018-05-31 | 2018-11-13 | 科大讯飞股份有限公司 | Information acquisition method and device |
| CN108804529A (en) * | 2018-05-02 | 2018-11-13 | 深圳智能思创科技有限公司 | A kind of question answering system implementation method based on Web |
| CN108921743A (en) * | 2018-06-20 | 2018-11-30 | 大国创新智能科技(东莞)有限公司 | Explain the puzzle method and education robot system of explaining the puzzle based on big data and artificial intelligence |
| CN108986910A (en) * | 2018-07-04 | 2018-12-11 | 平安科技（深圳）有限公司 | Answering method, device, computer equipment and storage medium on line |
| CN109002434A (en) * | 2018-05-31 | 2018-12-14 | 青岛理工大学 | Customer service question and answer matching method, server and storage medium |
| CN109241258A (en) * | 2018-08-23 | 2019-01-18 | 江苏索迩软件技术有限公司 | A kind of deep learning intelligent Answer System using tax field |
| CN109241266A (en) * | 2015-07-23 | 2019-01-18 | 上海智臻智能网络科技股份有限公司 | The creation method and apparatus that extension is asked are asked based on standard in human-computer interaction |
| CN109308320A (en) * | 2018-07-20 | 2019-02-05 | 北京智能点科技有限公司 | Conversation process configuration method is taken turns more by a kind of robot |
| WO2019062012A1 (en) * | 2017-09-30 | 2019-04-04 | 平安科技（深圳）有限公司 | Question association push method, electronic device and computer readable storage medium |
| CN109582798A (en) * | 2017-09-29 | 2019-04-05 | 阿里巴巴集团控股有限公司 | Automatic question-answering method, system and equipment |
| CN109583401A (en) * | 2018-12-05 | 2019-04-05 | 广东小天才科技有限公司 | Question searching method capable of automatically generating answers and user equipment |
| CN109657033A (en) * | 2018-12-13 | 2019-04-19 | 泰康保险集团股份有限公司 | Method, apparatus, storage medium and the electronic equipment of customer problem processing |
| CN109785698A (en) * | 2017-11-13 | 2019-05-21 | 上海流利说信息技术有限公司 | Method, apparatus, electronic equipment and medium for spoken language proficiency evaluation and test |
| CN109800286A (en) * | 2018-12-17 | 2019-05-24 | 北京百度网讯科技有限公司 | Talk with generation method and device |
| CN109829052A (en) * | 2019-02-19 | 2019-05-31 | 田中瑶 | A kind of open dialogue method and system based on human-computer interaction |
| CN109840277A (en) * | 2019-02-20 | 2019-06-04 | 西南科技大学 | A kind of government affairs Intelligent Service answering method and system |
| CN109885672A (en) * | 2019-03-04 | 2019-06-14 | 中国科学院软件研究所 | A question-and-answer intelligent retrieval system and method for online education |
| CN109918560A (en) * | 2019-01-09 | 2019-06-21 | 平安科技（深圳）有限公司 | A kind of answering method and device based on search engine |
| CN109977214A (en) * | 2019-03-29 | 2019-07-05 | 沈阳师范大学 | A kind of online knotty problem answer recommendation interactive approach and system of education of middle and primary schools |
| CN109992665A (en) * | 2019-03-14 | 2019-07-09 | 广州智语信息科技有限公司 | A kind of classification method based on the extension of problem target signature |
| CN110020009A (en) * | 2017-09-29 | 2019-07-16 | 阿里巴巴集团控股有限公司 | Online answering method, apparatus and system |
| CN110019706A (en) * | 2017-11-08 | 2019-07-16 | 北京搜狗科技发展有限公司 | A kind of problem generation method and device |
| CN110110133A (en) * | 2019-04-18 | 2019-08-09 | 贝壳技术有限公司 | A kind of intelligent sound data creation method and device |
| EP3531301A1 (en) * | 2018-02-27 | 2019-08-28 | DTMS GmbH | Computer-implemented method for querying data |
| CN110196901A (en) * | 2019-06-28 | 2019-09-03 | 北京百度网讯科技有限公司 | Construction method, device, computer equipment and the storage medium of conversational system |
| CN110263051A (en) * | 2019-06-11 | 2019-09-20 | 出门问问信息科技有限公司 | Question and answer for question answering system are to update method, device, equipment and storage medium |
| CN110263137A (en) * | 2019-05-31 | 2019-09-20 | 阿里巴巴集团控股有限公司 | The extracting method and device of subject key words, electronic equipment |
| CN110334272A (en) * | 2019-05-29 | 2019-10-15 | 平安科技（深圳）有限公司 | The intelligent answer method, apparatus and computer storage medium of knowledge based map |
| CN110347833A (en) * | 2019-07-09 | 2019-10-18 | 浙江工业大学 | A kind of classification method of more wheel dialogues |
| CN110472034A (en) * | 2019-08-21 | 2019-11-19 | 北京百度网讯科技有限公司 | Detection method, device, equipment and the computer readable storage medium of question answering system |
| CN110516057A (en) * | 2019-08-23 | 2019-11-29 | 深圳前海微众银行股份有限公司 | Method and device for answering petition questions |
| CN110516157A (en) * | 2019-08-30 | 2019-11-29 | 盈盛智创科技(广州)有限公司 | A kind of document retrieval method, equipment and storage medium |
| CN110659091A (en) * | 2018-06-29 | 2020-01-07 | 国际商业机器公司 | Session proxy dialog flow user interface |
| CN110955761A (en) * | 2019-10-12 | 2020-04-03 | 深圳壹账通智能科技有限公司 | Method and device for acquiring question and answer data in document, computer equipment and storage medium |
| CN111046155A (en) * | 2019-11-27 | 2020-04-21 | 中博信息技术研究院有限公司 | Semantic similarity calculation method based on FSM multi-turn question answering |
| CN111198940A (en) * | 2019-12-27 | 2020-05-26 | 北京百度网讯科技有限公司 | FAQ method, question-answer search system, electronic device, and storage medium |
| CN111428037A (en) * | 2020-03-24 | 2020-07-17 | 合肥科捷通科技信息服务有限公司 | Method for analyzing matching performance of behavior policy |
| CN111460095A (en) * | 2020-03-17 | 2020-07-28 | 北京百度网讯科技有限公司 | Question and answer processing method and device, electronic equipment and storage medium |
| CN111581329A (en) * | 2020-04-23 | 2020-08-25 | 上海兑观信息科技技术有限公司 | Short text matching method and device based on inverted index |
| CN111639164A (en) * | 2020-04-30 | 2020-09-08 | 中国平安财产保险股份有限公司 | Question-answer matching method and device of question-answer system, computer equipment and storage medium |
| CN111813889A (en) * | 2020-06-24 | 2020-10-23 | 北京安博盛赢教育科技有限责任公司 | Method, device, medium and electronic equipment for sorting question information |
| CN111930920A (en) * | 2020-09-30 | 2020-11-13 | 恒生电子股份有限公司 | Knowledge enhancement processing-based FAQ similarity calculation method and device and electronic equipment |
| CN112307291A (en) * | 2020-10-28 | 2021-02-02 | 福建菩泰网络科技有限公司 | Online consulting management method and management system for online shopping mall |
| CN112307171A (en) * | 2020-10-30 | 2021-02-02 | 中国电力科学研究院有限公司 | A system standard retrieval method and system based on electric power knowledge base and readable storage medium |
| CN112541109A (en) * | 2020-12-22 | 2021-03-23 | 北京百度网讯科技有限公司 | Answer abstract extraction method and device, electronic equipment, readable medium and product |
| CN112749265A (en) * | 2021-01-08 | 2021-05-04 | 哈尔滨工业大学 | Intelligent question-answering system based on multiple information sources |
| CN112925890A (en) * | 2021-03-05 | 2021-06-08 | 湖南神通智能股份有限公司 | Intelligent question-answering system |
| CN112948557A (en) * | 2021-03-09 | 2021-06-11 | 哈尔滨工业大学 | Iterative intelligent customer service system based on common problem set and assisted by description document |
| CN113836296A (en) * | 2021-09-28 | 2021-12-24 | 平安科技（深圳）有限公司 | Method, device, equipment and storage medium for generating Buddhist question-answer abstract |
| CN114219421A (en) * | 2021-11-13 | 2022-03-22 | 国网辽宁省电力有限公司葫芦岛供电公司 | An intelligent processing method of business requirements using big data |
| CN114491051A (en) * | 2022-04-02 | 2022-05-13 | 四川省大数据中心 | Project approval system for building site |
| CN116186232A (en) * | 2023-04-26 | 2023-05-30 | 中国电子技术标准化研究院 | Standard knowledge intelligent question-answering implementation method, device, equipment and medium |
| US11775839B2 (en) | 2020-06-10 | 2023-10-03 | International Business Machines Corporation | Frequently asked questions and document retrieval using bidirectional encoder representations from transformers (BERT) model trained on generated paraphrases |
| CN116881426A (en) * | 2023-08-30 | 2023-10-13 | 环球数科集团有限公司 | A self-explanatory question and answer system based on AIGC |
| CN118377887A (en) * | 2024-06-24 | 2024-07-23 | 阿里云飞天(杭州)云计算技术有限公司 | Automatic question and answer method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1928864A (en) * | 2006-09-22 | 2007-03-14 | 浙江大学 | FAQ based Chinese natural language ask and answer method |
| CN101097573A (en) * | 2006-06-28 | 2008-01-02 | 腾讯科技（深圳）有限公司 | Automatically request-answering system and method |
| CN101286161A (en) * | 2008-05-28 | 2008-10-15 | 华中科技大学 | A concept-based intelligent Chinese question answering system |
| CN101373532A (en) * | 2008-07-10 | 2009-02-25 | 昆明理工大学 | Implementation method of FAQ Chinese question answering system in the field of tourism |
| CN101377777A (en) * | 2007-09-03 | 2009-03-04 | 北京百问百答网络技术有限公司 | Automatic inquiring and answering method and system |
| CN103425640A (en) * | 2012-05-14 | 2013-12-04 | 华为技术有限公司 | Multimedia questioning-answering system and method |
-
2014
- 2014-02-27 CN CN201410068844.XA patent/CN103902652A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101097573A (en) * | 2006-06-28 | 2008-01-02 | 腾讯科技（深圳）有限公司 | Automatically request-answering system and method |
| CN1928864A (en) * | 2006-09-22 | 2007-03-14 | 浙江大学 | FAQ based Chinese natural language ask and answer method |
| CN101377777A (en) * | 2007-09-03 | 2009-03-04 | 北京百问百答网络技术有限公司 | Automatic inquiring and answering method and system |
| CN101286161A (en) * | 2008-05-28 | 2008-10-15 | 华中科技大学 | A concept-based intelligent Chinese question answering system |
| CN101373532A (en) * | 2008-07-10 | 2009-02-25 | 昆明理工大学 | Implementation method of FAQ Chinese question answering system in the field of tourism |
| CN103425640A (en) * | 2012-05-14 | 2013-12-04 | 华为技术有限公司 | Multimedia questioning-answering system and method |
Non-Patent Citations (3)
| Title |
|---|
| YONG JIANG ET AL.: ""Affiliation Disambiguation for Constructing Semantic Digital Libraries"", 《JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE AND TECHNOLOGY》 * |
| 刘里等: ""自动问答系统研究综述"", 《山东科技大学学报（自然科学版）》 * |
| 郑实福等: ""自动问答综述"", 《中文信息学报》 * |
Cited By (162)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102721A (en) * | 2014-07-18 | 2014-10-15 | 百度在线网络技术（北京）有限公司 | Method and device for recommending information |
| CN105335400A (en) * | 2014-07-22 | 2016-02-17 | 阿里巴巴集团控股有限公司 | Method and apparatus for obtaining answer information for questioning intention of user |
| CN105335400B (en) * | 2014-07-22 | 2018-11-23 | 阿里巴巴集团控股有限公司 | Enquirement for user is intended to obtain the method and device of answer information |
| CN104461525A (en) * | 2014-11-27 | 2015-03-25 | 韩慧健 | Intelligent user-defined consulting platform generating system |
| CN104461525B (en) * | 2014-11-27 | 2018-01-23 | 韩慧健 | A kind of intelligent consulting platform generation system that can customize |
| CN104679815A (en) * | 2014-12-08 | 2015-06-03 | 北京云知声信息技术有限公司 | Method and system for screening question and answer pairs and updating question and answer database in real time |
| CN104679815B (en) * | 2014-12-08 | 2018-02-23 | 北京云知声信息技术有限公司 | It is a kind of to screen question and answer pair and the method and system in real-time update question and answer storehouse |
| CN104536991A (en) * | 2014-12-10 | 2015-04-22 | 乐娟 | Answer extraction method and device |
| CN104536991B (en) * | 2014-12-10 | 2017-12-08 | 乐娟 | answer extracting method and device |
| CN107408110B (en) * | 2015-03-06 | 2020-12-15 | 国立研究开发法人情报通信研究机构 | Intentional pairing of expansion devices, recording media, and question-answering systems |
| CN107408110A (en) * | 2015-03-06 | 2017-11-28 | 国立研究开发法人情报通信研究机构 | Meaning pairing expansion device, computer program therefor, and question-response system |
| CN104778257B (en) * | 2015-04-20 | 2018-09-07 | 百度在线网络技术（北京）有限公司 | Using topic searching method and device |
| CN104778257A (en) * | 2015-04-20 | 2015-07-15 | 百度在线网络技术（北京）有限公司 | Method and device for searching applied problems |
| CN104820694B (en) * | 2015-04-28 | 2019-03-15 | 中国科学院自动化研究所 | Automatic question-answering method and system based on multiple knowledge base and integral linear programming ILP |
| CN104820694A (en) * | 2015-04-28 | 2015-08-05 | 中国科学院自动化研究所 | Automatic Q&A method and system based on multi-knowledge base and integral linear programming ILP |
| CN104951433A (en) * | 2015-06-24 | 2015-09-30 | 北京京东尚科信息技术有限公司 | Method and system for intention recognition based on context |
| CN104951433B (en) * | 2015-06-24 | 2018-01-23 | 北京京东尚科信息技术有限公司 | The method and system of intention assessment is carried out based on context |
| CN105095195B (en) * | 2015-07-03 | 2018-09-18 | 北京京东尚科信息技术有限公司 | Nan-machine interrogation's method and system of knowledge based collection of illustrative plates |
| CN105095195A (en) * | 2015-07-03 | 2015-11-25 | 北京京东尚科信息技术有限公司 | Method and system for human-machine questioning and answering based on knowledge graph |
| CN109241266A (en) * | 2015-07-23 | 2019-01-18 | 上海智臻智能网络科技股份有限公司 | The creation method and apparatus that extension is asked are asked based on standard in human-computer interaction |
| CN105354180A (en) * | 2015-08-26 | 2016-02-24 | 欧阳江 | Method and system for realizing open semantic interaction service |
| CN105117388A (en) * | 2015-09-21 | 2015-12-02 | 上海智臻智能网络科技股份有限公司 | Intelligent robot interaction system |
| CN105117387A (en) * | 2015-09-21 | 2015-12-02 | 上海智臻智能网络科技股份有限公司 | Intelligent robot interaction system |
| CN106547785B (en) * | 2015-09-22 | 2020-08-04 | 阿里巴巴集团控股有限公司 | Method and system for acquiring information in knowledge base |
| CN106547785A (en) * | 2015-09-22 | 2017-03-29 | 阿里巴巴集团控股有限公司 | Information getting method and system in knowledge base |
| CN105279274A (en) * | 2015-10-30 | 2016-01-27 | 北京京东尚科信息技术有限公司 | Answer combining and matching method and system based on natural synthetic answer system |
| CN105279274B (en) * | 2015-10-30 | 2018-11-02 | 北京京东尚科信息技术有限公司 | Answer synthesis based on naturally semantic question answering system and matched method and system |
| CN106649394A (en) * | 2015-11-03 | 2017-05-10 | 中兴通讯股份有限公司 | Fusion knowledge base processing method and device and knowledge base management system |
| WO2017076263A1 (en) * | 2015-11-03 | 2017-05-11 | 中兴通讯股份有限公司 | Method and device for integrating knowledge bases, knowledge base management system and storage medium |
| CN105574133A (en) * | 2015-12-15 | 2016-05-11 | 苏州贝多环保技术有限公司 | Multi-mode intelligent question answering system and method |
| CN105630887A (en) * | 2015-12-18 | 2016-06-01 | 北京中科汇联科技股份有限公司 | Representation method for knowledge markup languages of Chinese question answering system and Chinese question answering system |
| CN105740310B (en) * | 2015-12-21 | 2019-08-02 | 哈尔滨工业大学 | A kind of automatic answer method of abstracting and system in question answering system |
| CN105740310A (en) * | 2015-12-21 | 2016-07-06 | 哈尔滨工业大学 | Automatic answer summarizing method and system for question answering system |
| CN105630917A (en) * | 2015-12-22 | 2016-06-01 | 成都小多科技有限公司 | Intelligent answering method and intelligent answering device |
| CN105630938A (en) * | 2015-12-23 | 2016-06-01 | 深圳市智客网络科技有限公司 | Intelligent question-answering system |
| CN105653619A (en) * | 2015-12-25 | 2016-06-08 | 上海智臻智能网络科技股份有限公司 | Update method and device of correct log library in intelligent question-answering system |
| CN105653619B (en) * | 2015-12-25 | 2019-01-25 | 上海智臻智能网络科技股份有限公司 | The update method and device in correct log library in intelligent Answer System |
| CN106980624A (en) * | 2016-01-18 | 2017-07-25 | 阿里巴巴集团控股有限公司 | A kind for the treatment of method and apparatus of text data |
| CN105512349B (en) * | 2016-02-23 | 2019-03-26 | 首都师范大学 | A kind of answering method and device for learner's adaptive learning |
| CN105512349A (en) * | 2016-02-23 | 2016-04-20 | 首都师范大学 | Question and answer method and question and answer device for adaptive learning of learners |
| CN107168599B (en) * | 2016-03-07 | 2020-05-29 | 阿里巴巴集团控股有限公司 | Instant messaging method and device |
| CN107168599A (en) * | 2016-03-07 | 2017-09-15 | 阿里巴巴集团控股有限公司 | A kind of instant communicating method and device |
| CN107180037A (en) * | 2016-03-09 | 2017-09-19 | 北京京东尚科信息技术有限公司 | Man-machine interaction method and device |
| CN105824933B (en) * | 2016-03-18 | 2019-02-26 | 苏州大学 | Automatic question answering system based on main statement position and implementation method thereof |
| CN105824933A (en) * | 2016-03-18 | 2016-08-03 | 苏州大学 | Automatic question answering system based on main statement position and implementation method thereof |
| CN105893465A (en) * | 2016-03-28 | 2016-08-24 | 北京京东尚科信息技术有限公司 | Automatic question answering method and device |
| CN105893552A (en) * | 2016-03-31 | 2016-08-24 | 成都小多科技有限公司 | Method and device for processing data |
| CN105955976A (en) * | 2016-04-15 | 2016-09-21 | 中国工商银行股份有限公司 | Automatic answering system and method |
| CN105955976B (en) * | 2016-04-15 | 2019-05-14 | 中国工商银行股份有限公司 | A kind of automatic answering system and method |
| CN105912527A (en) * | 2016-04-19 | 2016-08-31 | 北京高地信息技术有限公司 | Method, device and system outputting answer according to natural language |
| CN105912697A (en) * | 2016-04-25 | 2016-08-31 | 北京光年无限科技有限公司 | Optimization method and device of dialog system knowledge base |
| CN105912697B (en) * | 2016-04-25 | 2019-08-27 | 北京光年无限科技有限公司 | A kind of optimization method and device of conversational system knowledge base |
| CN107341157B (en) * | 2016-04-29 | 2021-01-22 | 阿里巴巴集团控股有限公司 | Customer service conversation clustering method and device |
| CN107341157A (en) * | 2016-04-29 | 2017-11-10 | 阿里巴巴集团控股有限公司 | A kind of customer service dialogue clustering method and device |
| CN106095965A (en) * | 2016-06-17 | 2016-11-09 | 上海智臻智能网络科技股份有限公司 | A kind of data processing method and device |
| CN107528761A (en) * | 2016-06-21 | 2017-12-29 | 天脉聚源（北京）科技有限公司 | A kind of method and system of wechat public number automatic information reply |
| CN106202288A (en) * | 2016-06-30 | 2016-12-07 | 北京智能管家科技有限公司 | The optimization method of a kind of man-machine interactive system knowledge base and system |
| CN106202288B (en) * | 2016-06-30 | 2019-10-11 | 北京智能管家科技有限公司 | A kind of optimization method and system of man-machine interactive system knowledge base |
| CN108073664B (en) * | 2016-11-11 | 2021-08-31 | 北京搜狗科技发展有限公司 | Information processing method, device, equipment and client equipment |
| CN108073664A (en) * | 2016-11-11 | 2018-05-25 | 北京搜狗科技发展有限公司 | A kind of information processing method, device, equipment and client device |
| CN106599297A (en) * | 2016-12-28 | 2017-04-26 | 北京百度网讯科技有限公司 | Method and device for searching question-type search terms on basis of deep questions and answers |
| WO2018121759A1 (en) * | 2016-12-31 | 2018-07-05 | 深圳市优必选科技有限公司 | Intelligent question and answer method and system |
| WO2018133518A1 (en) * | 2017-01-20 | 2018-07-26 | 优酷网络技术（北京）有限公司 | Method, system, and apparatus for information interaction in live webcast |
| CN106878819A (en) * | 2017-01-20 | 2017-06-20 | 合网络技术（北京）有限公司 | The method, system and device of information exchange in a kind of network direct broadcasting |
| US11086912B2 (en) | 2017-03-03 | 2021-08-10 | Tencent Technology (Shenzhen) Company Limited | Automatic questioning and answering processing method and automatic questioning and answering system |
| CN106951470A (en) * | 2017-03-03 | 2017-07-14 | 中兴耀维科技江苏有限公司 | A kind of intelligent Answer System retrieved based on professional knowledge figure |
| WO2018157805A1 (en) * | 2017-03-03 | 2018-09-07 | 腾讯科技（深圳）有限公司 | Automatic questioning and answering processing method and automatic questioning and answering system |
| CN106934012A (en) * | 2017-03-10 | 2017-07-07 | 上海数眼科技发展有限公司 | A kind of question answering in natural language method and system of knowledge based collection of illustrative plates |
| CN106934012B (en) * | 2017-03-10 | 2020-05-08 | 上海数眼科技发展有限公司 | Natural language question-answering implementation method and system based on knowledge graph |
| CN106991161A (en) * | 2017-03-31 | 2017-07-28 | 北京字节跳动科技有限公司 | A kind of method for automatically generating open-ended question answer |
| CN107145510A (en) * | 2017-03-31 | 2017-09-08 | 西安科技大学 | A kind of mathematical formulae searching method and device |
| CN106991161B (en) * | 2017-03-31 | 2019-02-19 | 北京字节跳动科技有限公司 | A method of automatically generating open-ended question answer |
| CN107133210A (en) * | 2017-04-20 | 2017-09-05 | 中国科学院上海高等研究院 | Scheme document creation method and system |
| CN107229675A (en) * | 2017-04-28 | 2017-10-03 | 北京神州泰岳软件股份有限公司 | Question and answer base construction method, method of answering, the apparatus and system of list type knowledge |
| CN107301213A (en) * | 2017-06-09 | 2017-10-27 | 腾讯科技（深圳）有限公司 | Intelligent answer method and device |
| CN107436916A (en) * | 2017-06-15 | 2017-12-05 | 百度在线网络技术（北京）有限公司 | The method and device of intelligent prompt answer |
| CN107436916B (en) * | 2017-06-15 | 2021-04-27 | 百度在线网络技术（北京）有限公司 | Intelligent answer prompting method and device |
| CN107330040A (en) * | 2017-06-27 | 2017-11-07 | 李博 | One kind study topic searching method and its system |
| CN107453980A (en) * | 2017-07-26 | 2017-12-08 | 北京小米移动软件有限公司 | Problem response method and device in instant messaging |
| CN110020009B (en) * | 2017-09-29 | 2023-03-21 | 阿里巴巴集团控股有限公司 | Online question and answer method, device and system |
| CN109582798A (en) * | 2017-09-29 | 2019-04-05 | 阿里巴巴集团控股有限公司 | Automatic question-answering method, system and equipment |
| CN110020009A (en) * | 2017-09-29 | 2019-07-16 | 阿里巴巴集团控股有限公司 | Online answering method, apparatus and system |
| WO2019062012A1 (en) * | 2017-09-30 | 2019-04-04 | 平安科技（深圳）有限公司 | Question association push method, electronic device and computer readable storage medium |
| CN107729549A (en) * | 2017-10-31 | 2018-02-23 | 深圳追科技有限公司 | A kind of robot client service method and system comprising elements recognition |
| CN110019706A (en) * | 2017-11-08 | 2019-07-16 | 北京搜狗科技发展有限公司 | A kind of problem generation method and device |
| CN110019706B (en) * | 2017-11-08 | 2021-11-19 | 北京搜狗科技发展有限公司 | Question generation method and device |
| CN109785698A (en) * | 2017-11-13 | 2019-05-21 | 上海流利说信息技术有限公司 | Method, apparatus, electronic equipment and medium for spoken language proficiency evaluation and test |
| CN109785698B (en) * | 2017-11-13 | 2021-11-23 | 上海流利说信息技术有限公司 | Method, device, electronic equipment and medium for oral language level evaluation |
| CN108170780A (en) * | 2017-12-26 | 2018-06-15 | 北京邦邦共赢网络科技有限公司 | A kind of the problem of self-service question and answer matching process and device |
| CN108170679B (en) * | 2017-12-28 | 2021-09-03 | 中国联合网络通信集团有限公司 | Semantic matching method and system based on computer recognizable natural language description |
| CN108170679A (en) * | 2017-12-28 | 2018-06-15 | 中国联合网络通信集团有限公司 | It can recognize that the semantic matching method and system of natural language description based on computer |
| EP3531301A1 (en) * | 2018-02-27 | 2019-08-28 | DTMS GmbH | Computer-implemented method for querying data |
| CN108416033A (en) * | 2018-03-12 | 2018-08-17 | 北京汽车研究总院有限公司 | A kind of data analysing method, device and terminal for automobile research and development |
| CN108595494A (en) * | 2018-03-15 | 2018-09-28 | 腾讯科技（深圳）有限公司 | The acquisition methods and device of reply message |
| CN108595494B (en) * | 2018-03-15 | 2022-05-20 | 腾讯科技（深圳）有限公司 | Method and device for acquiring reply information |
| CN108804529A (en) * | 2018-05-02 | 2018-11-13 | 深圳智能思创科技有限公司 | A kind of question answering system implementation method based on Web |
| CN108595696A (en) * | 2018-05-09 | 2018-09-28 | 长沙学院 | A kind of human-computer interaction intelligent answering method and system based on cloud platform |
| CN108804627B (en) * | 2018-05-31 | 2021-04-06 | 科大讯飞股份有限公司 | Information acquisition method and device |
| CN108804627A (en) * | 2018-05-31 | 2018-11-13 | 科大讯飞股份有限公司 | Information acquisition method and device |
| CN108763529A (en) * | 2018-05-31 | 2018-11-06 | 苏州大学 | A kind of intelligent search method, device and computer readable storage medium |
| CN109002434A (en) * | 2018-05-31 | 2018-12-14 | 青岛理工大学 | Customer service question and answer matching method, server and storage medium |
| CN108921743B (en) * | 2018-06-20 | 2020-12-22 | 大国创新智能科技(东莞)有限公司 | Confusion method and confusion education robot system based on big data and artificial intelligence |
| CN108921743A (en) * | 2018-06-20 | 2018-11-30 | 大国创新智能科技(东莞)有限公司 | Explain the puzzle method and education robot system of explaining the puzzle based on big data and artificial intelligence |
| CN110659091A (en) * | 2018-06-29 | 2020-01-07 | 国际商业机器公司 | Session proxy dialog flow user interface |
| CN108986910A (en) * | 2018-07-04 | 2018-12-11 | 平安科技（深圳）有限公司 | Answering method, device, computer equipment and storage medium on line |
| CN108986910B (en) * | 2018-07-04 | 2023-09-05 | 平安科技（深圳）有限公司 | On-line question and answer method, device, computer equipment and storage medium |
| CN109308320A (en) * | 2018-07-20 | 2019-02-05 | 北京智能点科技有限公司 | Conversation process configuration method is taken turns more by a kind of robot |
| CN109241258A (en) * | 2018-08-23 | 2019-01-18 | 江苏索迩软件技术有限公司 | A kind of deep learning intelligent Answer System using tax field |
| CN109241258B (en) * | 2018-08-23 | 2022-09-06 | 江苏索迩软件技术有限公司 | Deep learning intelligent question-answering system applied to tax field |
| CN109583401A (en) * | 2018-12-05 | 2019-04-05 | 广东小天才科技有限公司 | Question searching method capable of automatically generating answers and user equipment |
| CN109657033A (en) * | 2018-12-13 | 2019-04-19 | 泰康保险集团股份有限公司 | Method, apparatus, storage medium and the electronic equipment of customer problem processing |
| CN109800286A (en) * | 2018-12-17 | 2019-05-24 | 北京百度网讯科技有限公司 | Talk with generation method and device |
| CN109918560B (en) * | 2019-01-09 | 2024-03-12 | 平安科技（深圳）有限公司 | Question and answer method and device based on search engine |
| CN109918560A (en) * | 2019-01-09 | 2019-06-21 | 平安科技（深圳）有限公司 | A kind of answering method and device based on search engine |
| CN109829052A (en) * | 2019-02-19 | 2019-05-31 | 田中瑶 | A kind of open dialogue method and system based on human-computer interaction |
| CN109840277A (en) * | 2019-02-20 | 2019-06-04 | 西南科技大学 | A kind of government affairs Intelligent Service answering method and system |
| CN109885672A (en) * | 2019-03-04 | 2019-06-14 | 中国科学院软件研究所 | A question-and-answer intelligent retrieval system and method for online education |
| CN109885672B (en) * | 2019-03-04 | 2020-10-30 | 中国科学院软件研究所 | Question-answering type intelligent retrieval system and method for online education |
| CN109992665A (en) * | 2019-03-14 | 2019-07-09 | 广州智语信息科技有限公司 | A kind of classification method based on the extension of problem target signature |
| CN109977214A (en) * | 2019-03-29 | 2019-07-05 | 沈阳师范大学 | A kind of online knotty problem answer recommendation interactive approach and system of education of middle and primary schools |
| CN110110133A (en) * | 2019-04-18 | 2019-08-09 | 贝壳技术有限公司 | A kind of intelligent sound data creation method and device |
| CN110334272B (en) * | 2019-05-29 | 2022-04-12 | 平安科技（深圳）有限公司 | Intelligent question-answering method and device based on knowledge graph and computer storage medium |
| CN110334272A (en) * | 2019-05-29 | 2019-10-15 | 平安科技（深圳）有限公司 | The intelligent answer method, apparatus and computer storage medium of knowledge based map |
| CN110263137A (en) * | 2019-05-31 | 2019-09-20 | 阿里巴巴集团控股有限公司 | The extracting method and device of subject key words, electronic equipment |
| CN110263137B (en) * | 2019-05-31 | 2023-06-06 | 创新先进技术有限公司 | Theme keyword extraction method and device and electronic equipment |
| CN110263051A (en) * | 2019-06-11 | 2019-09-20 | 出门问问信息科技有限公司 | Question and answer for question answering system are to update method, device, equipment and storage medium |
| CN110196901A (en) * | 2019-06-28 | 2019-09-03 | 北京百度网讯科技有限公司 | Construction method, device, computer equipment and the storage medium of conversational system |
| CN110347833A (en) * | 2019-07-09 | 2019-10-18 | 浙江工业大学 | A kind of classification method of more wheel dialogues |
| CN110472034A (en) * | 2019-08-21 | 2019-11-19 | 北京百度网讯科技有限公司 | Detection method, device, equipment and the computer readable storage medium of question answering system |
| CN110472034B (en) * | 2019-08-21 | 2022-11-15 | 北京百度网讯科技有限公司 | Detection method, device and equipment of question-answering system and computer readable storage medium |
| CN110516057B (en) * | 2019-08-23 | 2022-10-28 | 深圳前海微众银行股份有限公司 | Petition question answering method and device |
| CN110516057A (en) * | 2019-08-23 | 2019-11-29 | 深圳前海微众银行股份有限公司 | Method and device for answering petition questions |
| CN110516157A (en) * | 2019-08-30 | 2019-11-29 | 盈盛智创科技(广州)有限公司 | A kind of document retrieval method, equipment and storage medium |
| CN110955761A (en) * | 2019-10-12 | 2020-04-03 | 深圳壹账通智能科技有限公司 | Method and device for acquiring question and answer data in document, computer equipment and storage medium |
| WO2021068615A1 (en) * | 2019-10-12 | 2021-04-15 | 深圳壹账通智能科技有限公司 | Method and device for acquiring question and answer data in document, computer device, and storage medium |
| CN111046155A (en) * | 2019-11-27 | 2020-04-21 | 中博信息技术研究院有限公司 | Semantic similarity calculation method based on FSM multi-turn question answering |
| CN111198940A (en) * | 2019-12-27 | 2020-05-26 | 北京百度网讯科技有限公司 | FAQ method, question-answer search system, electronic device, and storage medium |
| CN111198940B (en) * | 2019-12-27 | 2023-01-31 | 北京百度网讯科技有限公司 | FAQ method, question-answer search system, electronic device, and storage medium |
| CN111460095A (en) * | 2020-03-17 | 2020-07-28 | 北京百度网讯科技有限公司 | Question and answer processing method and device, electronic equipment and storage medium |
| CN111428037B (en) * | 2020-03-24 | 2022-09-20 | 合肥科捷通科技信息服务有限公司 | Method for analyzing matching performance of behavior policy |
| CN111428037A (en) * | 2020-03-24 | 2020-07-17 | 合肥科捷通科技信息服务有限公司 | Method for analyzing matching performance of behavior policy |
| CN111581329A (en) * | 2020-04-23 | 2020-08-25 | 上海兑观信息科技技术有限公司 | Short text matching method and device based on inverted index |
| CN111639164A (en) * | 2020-04-30 | 2020-09-08 | 中国平安财产保险股份有限公司 | Question-answer matching method and device of question-answer system, computer equipment and storage medium |
| US11775839B2 (en) | 2020-06-10 | 2023-10-03 | International Business Machines Corporation | Frequently asked questions and document retrieval using bidirectional encoder representations from transformers (BERT) model trained on generated paraphrases |
| CN111813889A (en) * | 2020-06-24 | 2020-10-23 | 北京安博盛赢教育科技有限责任公司 | Method, device, medium and electronic equipment for sorting question information |
| CN111930920A (en) * | 2020-09-30 | 2020-11-13 | 恒生电子股份有限公司 | Knowledge enhancement processing-based FAQ similarity calculation method and device and electronic equipment |
| CN112307291A (en) * | 2020-10-28 | 2021-02-02 | 福建菩泰网络科技有限公司 | Online consulting management method and management system for online shopping mall |
| CN112307171A (en) * | 2020-10-30 | 2021-02-02 | 中国电力科学研究院有限公司 | A system standard retrieval method and system based on electric power knowledge base and readable storage medium |
| CN112307171B (en) * | 2020-10-30 | 2022-02-11 | 中国电力科学研究院有限公司 | A system standard retrieval method and system based on electric power knowledge base and readable storage medium |
| CN112541109A (en) * | 2020-12-22 | 2021-03-23 | 北京百度网讯科技有限公司 | Answer abstract extraction method and device, electronic equipment, readable medium and product |
| CN112541109B (en) * | 2020-12-22 | 2023-10-24 | 北京百度网讯科技有限公司 | Answer abstract extraction method and device, electronic equipment, readable medium and product |
| CN112749265A (en) * | 2021-01-08 | 2021-05-04 | 哈尔滨工业大学 | Intelligent question-answering system based on multiple information sources |
| CN112925890A (en) * | 2021-03-05 | 2021-06-08 | 湖南神通智能股份有限公司 | Intelligent question-answering system |
| CN112948557A (en) * | 2021-03-09 | 2021-06-11 | 哈尔滨工业大学 | Iterative intelligent customer service system based on common problem set and assisted by description document |
| CN113836296A (en) * | 2021-09-28 | 2021-12-24 | 平安科技（深圳）有限公司 | Method, device, equipment and storage medium for generating Buddhist question-answer abstract |
| CN114219421A (en) * | 2021-11-13 | 2022-03-22 | 国网辽宁省电力有限公司葫芦岛供电公司 | An intelligent processing method of business requirements using big data |
| CN114491051B (en) * | 2022-04-02 | 2022-07-29 | 四川省大数据中心 | Project approval system for building site |
| CN114491051A (en) * | 2022-04-02 | 2022-05-13 | 四川省大数据中心 | Project approval system for building site |
| CN116186232A (en) * | 2023-04-26 | 2023-05-30 | 中国电子技术标准化研究院 | Standard knowledge intelligent question-answering implementation method, device, equipment and medium |
| CN116881426A (en) * | 2023-08-30 | 2023-10-13 | 环球数科集团有限公司 | A self-explanatory question and answer system based on AIGC |
| CN116881426B (en) * | 2023-08-30 | 2023-11-10 | 环球数科集团有限公司 | A self-explanatory question and answer system based on AIGC |
| CN118377887A (en) * | 2024-06-24 | 2024-07-23 | 阿里云飞天(杭州)云计算技术有限公司 | Automatic question and answer method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103902652A (en) | Automatic question-answering system | |
| Mihalcea et al. | Wikify! Linking documents to encyclopedic knowledge | |
| Cafarella et al. | Webtables: exploring the power of tables on the web | |
| CN102087669B (en) | Intelligent search engine system based on semantic association | |
| US9715493B2 (en) | Method and system for monitoring social media and analyzing text to automate classification of user posts using a facet based relevance assessment model | |
| Cafarella et al. | Web-scale extraction of structured data | |
| Lin et al. | An integrated approach to extracting ontological structures from folksonomies | |
| CN105868313A (en) | Mapping knowledge domain questioning and answering system and method based on template matching technique | |
| Lytvyn et al. | Identifying textual content based on thematic analysis of similar texts in big data | |
| CN114090861A (en) | Education field search engine construction method based on knowledge graph | |
| Cafarella et al. | Relational web search | |
| Nabil et al. | AlQuAnS-An Arabic Language Question Answering System. | |
| CN105677684A (en) | Method for making semantic annotations on content generated by users based on external data sources | |
| Wroblewski | A hierarchical www pages clustering algorithm based on the vector space model | |
| Gupta et al. | Document summarisation based on sentence ranking using vector space model | |
| Kim et al. | Ontology construction using online ontologies based on selection, mapping and merging | |
| Chen et al. | FAQ system in specific domain based on concept hierarchy and question type | |
| Park et al. | Extracting search intentions from web search logs | |
| Sahu et al. | Analytical study on intelligent information retrieval system using semantic network | |
| Sahu et al. | Review of web document clustering algorithms | |
| Takhirov et al. | An evidence-based verification approach to extract entities and relations for knowledge base population | |
| Ibekwe‐SanJuan | Information Science in the web era: A term‐based approach to domain mapping | |
| TWI423053B (en) | Domain Interpretation Data Retrieval Method and Its System | |
| Sharma et al. | Improved stemming approach used for text processing in information retrieval system | |
| Kanavos et al. | Extracting knowledge from web search engine using wikipedia |
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 |
Application publication date: 20140702 |
|
| RJ01 | Rejection of invention patent application after publication |