CN102236677A - Question answering system-based information matching method and system - Google Patents
Question answering system-based information matching method and system Download PDFInfo
- Publication number
- CN102236677A CN102236677A CN2010101662754A CN201010166275A CN102236677A CN 102236677 A CN102236677 A CN 102236677A CN 2010101662754 A CN2010101662754 A CN 2010101662754A CN 201010166275 A CN201010166275 A CN 201010166275A CN 102236677 A CN102236677 A CN 102236677A
- Authority
- CN
- China
- Prior art keywords
- question
- information
- user
- user interaction
- interaction behavior
- 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
- 238000000034 method Methods 0.000 title claims abstract description 38
- 230000003993 interaction Effects 0.000 claims abstract description 78
- 238000012512 characterization method Methods 0.000 claims abstract description 54
- 230000004044 response Effects 0.000 claims abstract description 9
- 238000004458 analytical method Methods 0.000 claims description 11
- 238000005516 engineering process Methods 0.000 claims description 10
- 230000009467 reduction Effects 0.000 claims description 6
- 230000008878 coupling Effects 0.000 description 8
- 238000010168 coupling process Methods 0.000 description 8
- 238000005859 coupling reaction Methods 0.000 description 8
- 238000012423 maintenance Methods 0.000 description 8
- 238000011156 evaluation Methods 0.000 description 7
- 238000012163 sequencing technique Methods 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 235000012364 Peperomia pellucida Nutrition 0.000 description 1
- 240000007711 Peperomia pellucida Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 230000003412 degenerative effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000013011 mating Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
本发明公开了一种基于问答系统的信息匹配方法及系统，所述方法包括：为问答系统知识库中的问答对配置用户交互行为表征参数，并根据用户交互行为反馈信息对所述用户交互行为表征参数进行动态更新，所述用户交互行为反馈信息为对问答系统的提问信息响应的反馈；根据用户输入的提问信息在问答系统知识库中获取问题候选集，以所述问题候选集中的各条问题信息的用户交互行为表征参数为索引，对所述问题候选集中的各条问题信息进行排序，向用户返回排序在前的预定数量的问题信息。本发明能够提高问答系统的信息匹配速度和精度。
The invention discloses an information matching method and system based on a question answering system. The method includes: configuring user interaction behavior characterization parameters for question answer pairs in the question answering system knowledge base, and analyzing the user interaction behavior according to user interaction behavior feedback information. The characterization parameters are dynamically updated, and the feedback information of the user interaction behavior is the feedback of the response to the question information of the question answering system; the question candidate set is obtained in the question answer system knowledge base according to the question information input by the user, and each item in the question candidate set The user interaction behavior characterization parameter of the question information is an index, sorts each piece of question information in the question candidate set, and returns a predetermined number of question information ranked first to the user. The invention can improve the information matching speed and accuracy of the question answering system.
Description
Technical field
The present invention relates to search engine technique, specifically, relate to a kind of information matching method and system based on question answering system.
Background technology
There are some researches show, in the face of the mass users information of internet, most of users all with search engine as access internet to obtain the starting point of information.Traditional search engines adopts the mode of keyword matching to return the web page listings that comprises institute's searching key word, by checking after user's webpage clicking based on Webpage search mostly.Along with the continuous development of science and technology, mobile phone, PDA handheld devices such as (personal digital assistants) are popularized gradually, and increasing user is by the handheld device access internet, and the drawback of traditional search engines also begins to manifest.The screen size of handheld device and communication efficiency be to the restriction of information capacity, and the web page listings that makes traditional search engines return is not suitable for and shows the user.Question answering system is one of scheme that solves this contradiction.
The difference of question answering system and conventional web search is that the user can put question to system with the sentence of natural language form, behind the systematic analysis customer problem, returns the answer of customer problem accurately.For example, the user wishes search problem: " which bar the longest river of China is? " conventional web search is returned may be the web page address that comprises key words such as " China ", " the longest ", " river ", " where ", oneself searches answer after the user needs webpage clicking.And utilize question answering system, will directly return user's answer " the Changjiang river ".Avoid secondary to browse, improved user experience.
Which kind of method is present question answering system no matter adopt carry out the answer retrieval, and big city is equipped with the question and answer knowledge base, and this knowledge base is often stored some common problem, and the answer of problem is adopted to climb automatically and got or mode that the expert safeguards obtains.When the user puts question to,, then directly return the answer in the knowledge base if the match is successful for the problem in problem and the knowledge base.The structure of knowledge base mostly is man-to-man question and answer to form, promptly for example: problem is " which bar the longest river of China is ", answer is " the Changjiang river ", and one group of such question and answer are right as one group of question and answer in the knowledge base, and the question and answer that store the such frequently asked question of many groups in the knowledge base are right.Question answering system is searched for the problem form in the knowledge base of mating most with user's enquirement in knowledge base according to user's enquirement, and then provide the answer of this problem.
In question sentence matching process based on knowledge base, go up from sorting, existent method concentrates on three aspects such as semanteme, grammer and conventional information opinion sort method.Traditional information theory sort method, limited as TF-IDF value sort method commonly used in Webpage search effect in the question sentence coupling.Because the question sentence of question answering system is often shorter, word frequency information is relatively concentrated also fewer in question sentence, is mostly 1-2 time, therefore is difficult to embody the meaning that TF-IDF is comprised on word frequency information.
Utilize the general needs traversal of sort method (comprising the vector space method of the utilizing the semantic information computed range) Candidate Set of semantic matches degree, the problem that the life period complexity is high.And semantic distance information quality so far is not high, and reliability and spreadability have limited the scope of application.
Sort method based on the syntactic information matching degree, because the dirigibility of Chinese and grammer be severity not, replenishing preceding two kinds of methods often, extract 1-2 tangible grammar property and be used for calculating ordering, can not be used for sort method separately, and there is the not high problem of calculation of complex and precision equally, can directly influences ranking results.
To sum up, in the prior art, the information matches in the question answering system, specifically, still there are many defectives in the matching process of question sentence information, and matching precision and matching efficiency all have much room for improvement.
Summary of the invention
Technical matters to be solved by this invention is, a kind of information matching method and system based on question answering system is provided, and realization that can efficiently and accurately is based on the information matches of question answering system.
For solving the problems of the technologies described above, the present invention has adopted following technical scheme:
A kind of information matching method based on question answering system comprises:
For the question and answer in the question answering system knowledge base to configure user interbehavior characterization parameter, and according to user interactions behavior feedback information described user interactions behavior characterization parameter is dynamically updated, described user interactions behavior feedback information is the feedback to the enquirement information response of question answering system;
Enquirement information according to user's input is obtained the problem Candidate Set in the question answering system knowledge base, user interactions behavior characterization parameter with each the bar problem information in the described problem Candidate Set is an index, each bar problem information in the described problem Candidate Set is sorted, return the problem information of predetermined quantity the preceding that sorts to the user.
In an embodiment of the present invention, upgrade described user interactions behavior characterization parameter in the following manner:, improve user interactions behavior characterization parameter if user interactions behavior feedback information is positive feedback; If user interactions behavior feedback information is negative feedback, reduce user interactions behavior characterization parameter.
In an embodiment of the present invention, also comprise: penalty factor is set, with the reduction degree of weighting user interactions behavior feedback information to user interactions behavior characterization parameter.
In an embodiment of the present invention, obtaining the problem Candidate Set according to the enquirement information of user's input in the question answering system knowledge base carries out in the following way:
Utilize the natural language technology that user's information of puing question to is carried out grammatical analysis, from user's enquirement information, extract keyword according to grammatical analysis;
Retrieval comprises described crucial word problem information in the question answering system knowledge base, according to default keyword coverage rate threshold value, the problem information that meets or exceeds described keyword coverage rate threshold value is selected the described problem Candidate Set of formation.
In an embodiment of the present invention, described user interactions behavior characterization parameter is the product of global frequencies and keyword coverage rate.
In an embodiment of the present invention, described user interactions behavior characterization parameter is by webpage, WAP page or SMS transmission.
The present invention also provides a kind of information matches system based on question answering system, comprising:
Characterization parameter is provided with module, be used to question and answer in the question answering system knowledge base to configure user interbehavior characterization parameter, and according to user interactions behavior feedback information described user interactions behavior characterization parameter is dynamically updated, described user interactions behavior feedback information is the feedback to the enquirement information response of question answering system;
The information matches module, be used for obtaining the problem Candidate Set in the question answering system knowledge base according to the enquirement information of user's input, user interactions behavior characterization parameter with each the bar problem information in the described problem Candidate Set is an index, each bar problem information in the described problem Candidate Set is sorted, return the problem information of predetermined quantity the preceding that sorts to the user.
Beneficial effect of the present invention is:
By being that question and answer in the question answering system knowledge base are to configure user interbehavior characterization parameter, enquirement information according to user's input is obtained the problem Candidate Set in the question answering system knowledge base, and be index with user interactions behavior characterization parameter, each bar problem information in the described problem Candidate Set is sorted, return the problem information of predetermined quantity the preceding that sorts to the user, because user interactions behavior characterization parameter dynamically updates according to user interactions behavior feedback information, can constantly revise match information in real time, the accuracy of the match information that assurance is returned to the user, and only return the ordering information of predetermined quantity the preceding in the problem Candidate Set, thereby can realize information matches efficiently based on question answering system.
Description of drawings
Fig. 1 is the method flow diagram of an embodiment of the present invention;
Fig. 2 is the system works schematic diagram of an embodiment of the present invention.
Embodiment
In conjunction with the accompanying drawings the present invention is described in further detail below by embodiment.
As shown in Figure 1, the information matching method based on question answering system of the present invention mainly comprises:
For the question and answer in the question answering system knowledge base to configure user interbehavior characterization parameter, and according to user interactions behavior feedback information described user interactions behavior characterization parameter is dynamically updated, described user interactions behavior feedback information is the feedback to the enquirement information response of question answering system;
Enquirement information according to user's input is obtained the problem Candidate Set in the question answering system knowledge base, user interactions behavior characterization parameter with each the bar problem information in the described problem Candidate Set is an index, each bar problem information in the described problem Candidate Set is sorted, return the problem information of predetermined quantity the preceding that sorts to the user.
In order to improve the accuracy of information search and coupling, general, can in the question answering system knowledge base, obtain the problem Candidate Set according to the enquirement information of user's input.The problem Candidate Set, promptly system is at user's enquirement behavior and a plurality of similar problem that searches from knowledge base.
" problem in problem of being retrieved and the question answering system knowledge base is the probability of same problem " is defined as matching degree, thereby a plurality of similar problem among the problem candidate all has its matching degree separately, the matching degree determination methods for example can adopt TF-IDF (term frequency-inverse documentfrequency, be used for the weighting technique that information retrieval and information are prospected) method, perhaps judge by calculating the keyword coverage rate.
Keyword coverage rate for problem, normally put question to the question sentence of behavior to carry out grammatical analysis to the user, for example for " China the longest river be any bar? " question sentence, by grammatical analysis, extracting keyword for example comprises: China, the longest, river, and seek again and include these crucial word problems in the knowledge base.Problem in the knowledge base, the keyword of covering is many more, shows that then matching degree is high more.
Since the question sentence that the user puts question to behavior not necessarily with knowledge base in the problem form of storing in full accord, by obtaining the mode of problem Candidate Set, can avoid omission, promptly guarantee to have in the knowledge base associated answer but in matching process, to be excluded because problem form and user put question to different.Thereby can guarantee information the validity and the accuracy of coupling.
The problem information (because the question and answer in the question answering system also can be answer information to form) of predetermined quantity in the Candidate Set (can for example be 1 by default) can be returned to the user.General, for raising the efficiency (by reducing the data volume returned to the user) to improve transfer efficiency and system overhead conserved, also may be subjected to the user to use handheld device to limit, only be one or part bar information in the problem Candidate Set to the problem information that the user returns.For this reason, need decision return which bar or which bar information to the user, this can be by to the problem information sorting in the problem Candidate Set and return and sort the preceding that the problem information of predetermined quantity is solved.The index of ordering can adopt matching degree order from high to low, yet this order might not really reflect the coupling of information.For improving the matching precision of information, the information matching method of the embodiment of the invention, adopted and be the mode of the question and answer in the question answering system knowledge base configure user interbehavior characterization parameter, this user interactions behavior can dynamically update according to user interactions behavior feedback information, for example, dynamically update and to carry out in the following manner:, improve user interactions behavior characterization parameter if user interactions behavior feedback information is positive feedback; If user interactions behavior feedback information is negative feedback, reduce user interactions behavior characterization parameter.
Penalty factor can be set, with the reduction degree of weighting user interactions behavior feedback information to user interactions behavior characterization parameter.Penalty factor mainly is based on the actual use pattern of user and considers, in enforcement of the present invention, the user interactions behavior mainly comprises two kinds, a kind of enquirement behavior that is the user to question answering system, a kind of evaluation behavior that is the user to the enquirement information response of question answering system, being question answering system responds at user's enquirement behavior, Candidate Set and final problem from predetermined quantity to the user or the answer information of returning have problems, the user can estimate problem or answer information that question answering system is returned, if think coupling, can estimate " satisfaction ",, can estimate " being unsatisfied with " if think and do not match.For user's enquirement behavior, the feedback information that then is considered as the user interactions behavior is positive feedback, then corresponding raising user interactions behavior characterization parameter; For the evaluation of user behavior, the evaluation behavior of " satisfaction " also is considered as positive feedback; And the evaluation behavior of " being unsatisfied with ", then be considered as negative feedback, then want corresponding reduction user interactions behavior characterization parameter, yet, consider that problem or answer information that the user does not always return question answering system estimate, and positive feedback has produced in user's enquirement behavior, thereby need to a certain degree amplify degenerative influence by penalty factor.
In an embodiment of the present invention, obtaining the problem Candidate Set according to the enquirement information of user's input in the question answering system knowledge base can carry out in the following way:
Utilize the natural language technology that user's information of puing question to is carried out grammatical analysis, from user's enquirement information, extract keyword according to grammatical analysis; The natural language processing technique that native system relates to comprises technology such as participle and part-of-speech tagging.The natural language technology is widely used in being not described in detail in this in the middle of the problem coupling and answer ordering of question answering system.
Retrieval comprises described crucial word problem information in the question answering system knowledge base, according to default keyword coverage rate threshold value, the problem information that meets or exceeds described keyword coverage rate threshold value is selected the described problem Candidate Set of formation.
The height of keyword coverage rate threshold value has determined the balance between loss and matching efficiency, and this value is high more, and then problem Candidate Set scope is more little, and the possibility of omission is then big more; This value is low more, and then problem Candidate Set scope is big more, the corresponding reduction of matching efficiency.Thereby can be by experiment, emulation or The actual running results set a comparatively appropriate threshold.
User interactions behavior characterization parameter can pass through webpage, WAP page or SMS transmission.Be that the user can pass through handheld device log-on webpage or WAP page, at webpage or WAP page is putd question to or the evaluation behavior, perhaps put question to or the evaluation behavior the corresponding user interactions behavior characterization parameter that obtains of question answering system by way of short messages.
The information matches system based on question answering system of the embodiment of the invention comprises:
Characterization parameter is provided with module, be used to question and answer in the question answering system knowledge base to configure user interbehavior characterization parameter, and according to user interactions behavior feedback information described user interactions behavior characterization parameter is dynamically updated, described user interactions behavior feedback information is the feedback to the enquirement information response of question answering system;
The information matches module, be used for obtaining the problem Candidate Set in the question answering system knowledge base according to the enquirement information of user's input, user interactions behavior characterization parameter with each the bar problem information in the described problem Candidate Set is an index, each bar problem information in the described problem Candidate Set is sorted, return the problem information of predetermined quantity the preceding that sorts to the user.
This information matches system, suitable, can adopt software, hardware or software to add hardware mode and realize, for example, in application server, carry out the computer executable program of the above-mentioned information matches function of realization etc.
The present invention by user interactions behavior characterization parameter setting and dynamically update, precision and speed that can the guarantee information coupling realize high efficiency and accuracy based on the information matches of question answering system.
As shown in Figure 2, in an application examples of the present invention, information matches system based on question answering system comprises frequency collector unit, knowledge base maintenance unit, sequencing unit, retrieval unit, be used between terminal user and question answering system knowledge base, realizing information matches, wherein, the function of information matches module is above mainly realized by sequencing unit and retrieval unit; The function that characterization parameter above is provided with module is mainly realized by frequency collector unit and knowledge base maintenance unit.That is to say that in this application examples, user interactions behavior characterization parameter is a frequency information.
The frequency collector unit is the unit with user interactions, can webpage, the form and the user of WAP page or SMS carry out alternately.When the user asks a question to question answering system, collect positive frequency, when the user is unsatisfied with information to the problem transmission of being mated, collect negative frequency.
The knowledge base maintenance unit is according to the matching degree of customer problem (user puts question to the problem form of behavior) and knowledge base problem (the problem form of storing in the knowledge base), and positive frequency or negative frequency information that the frequency collector unit is collected are updated in the knowledge base in proportion.Utilize two factors of matching degree and frequency to carry out integrated ordered in the search problem sequencer procedure to problem.
The principle of work of said system is: when terminal user and question answering system were mutual, the frequency collector unit was selected to collect positive frequency or negative frequency according to user's interbehavior, if the enquirement behavior then sends retrieval request to retrieval unit and obtains the problem Candidate Set.
If the user interactions behavior is that the problem that system returns is provided the evaluation behavior of whether being satisfied with, then at this single problem of being estimated, by the frequency maintenance unit with the frequency shift information updating of problem Candidate Set in knowledge base, simultaneously, the problem Candidate Set that retrieval unit is retrieved sorts jointly by matching degree and frequency, and extract and sort in last position or several problem or answer, return to the user.
The frequency collector unit is as follows to the treatment step of user's input:
Step S1): beginning;
Step S2): receive user's input;
Step S3): judge user's input,, collect positive frequency if input is search problem, if the negative input of the match information that question answering system is returned, the negative frequency of collecting this customizing messages.Wherein, the step of the collection positive frequency among the step S3 comprises:
Step S311): the retrieval unit that customer problem is sent to question answering system;
Step S312): utilize retrieval unit from knowledge base, to return a problem Candidate Set;
Step S313): the problem Candidate Set by the ordering of problem matching degree, is chosen the problem of matching degree more than N (N＞0) in the problem Candidate Set, give the positive integer frequency in proportion.
The size of N and the frequency ratio of giving all is a constant, and can adjust according to system's operating position.For example, set N=90%, comprise in the problem Candidate Set that promptly matching degree is in the problem more than 90%.Give ratio for frequency, for example the problem that matching degree can be reached more than 90% is given positive frequency 9; Give positive frequency 8 with matching degree in the problem of 80-90%.
Among the above-mentioned steps S312, retrieval unit can adopt following steps that knowledge base is retrieved:
Step J1): utilize the natural language technology that customer problem is carried out participle and part-of-speech tagging;
Step J2): from problem, select keyword according to part of speech and syntactic structure;
Step J3): in the inverted file of knowledge base, search successively and comprise crucial word problem;
Step J4): return the keyword percentage of coverage problem more than M at least.
As mentioned before, the Candidate Set problem matching degree described in the step S313 can be judged with the keyword coverage rate (being the keyword percentage of coverage) of problem content.
The user can be provided with button click in question answering system in the result that the user returns shows for convenience, is used for carrying out the negative input of the user described in the step S3, if the user thinks that this result is not the result who oneself wishes, can click this button.
The method of the collection negative frequency described in the step S3 is meant at this particular problem, at first by giving the positive integer frequency f with the matching degree of customer problem.For example the user puts question to that sentence is " which bar the longest river of China is? ", and the problem of storing in the knowledge base be " China the longest river be any bar? ", matching degree reaches more than 90%, and the positive frequency of for example giving the problem of knowledge base storage this moment is 9.
Suppose that in the problem Candidate Set this problem matching degree is the highest, thereby return the corresponding answer of this problem or this problem, but user feedback is dissatisfied to this problem returned or answer, then for example multiply by penalty factor-1, obtains negative frequency-9 to the user.Penalty factor is a constant in system herein.Because in user search problem (user's enquirement behavior), frequency f has been updated in the knowledge base, if the user thinks that problem does not match, should cancel the frequency of current renewal so, but use pattern based on the user, the user may not necessarily click " being unsatisfied with " at every turn, that is to say, inappropriate positive frequency some the time by user " being unsatisfied with " feedback deduct; Some the time then because user's feedback and not being eliminated not.The influence that therefore should have a penalty factor to be used for amplifying " be unsatisfied with ", i.e. the overall situation, if think that the user at every turn can be to the click of unsatisfied problem response " being unsatisfied with ", then penalty factor is-1; And if hypothesis has only 10% dissatisfied meeting clicked, penalty factor should be-10 so.So the real figure of penalty factor should be adjusted according to the system actual operating state, use by the positive integer frequency f of giving with the matching degree of customer problem and multiply by penalty factor K (K＜0), obtain a negative frequency.
Wherein, the ratio of giving of the size of K and f can be adjusted to some extent according to system's operating position.
The plus or minus frequency of the different problems that the frequency maintenance unit will be collected is updated in the knowledge base.If knowledge base itself does not contain frequency information and content can not be changed, then can set up and knowledge base problem frequency record storehouse one to one, record each question and answer in the knowledge base to and the corresponding relation of frequency information, frequency information is updated in the frequency record storehouse.
Can utilize database to come the content and the frequency information in stored knowledge storehouse, then the frequency maintenance unit will be the renewal processing unit that is connected with this database.
The job step of the sequencing unit in Fig. 2 example is as follows:
Step P1): beginning;
Step P2): receive the problem Candidate Set that question answering system is returned;
Step P3): the frequency information that from knowledge base or frequency record storehouse, obtains the problem Candidate Set;
Step P4): utilize the matching degree of each problem in the problem Candidate Set and frequency information that problem is sorted;
Step P5):, from knowledge base, obtain corresponding answer to the problem of ordering in the first place;
Step P6): answer is showed the user.
Question answering system can use WAP site as exhibition method, and the user imports the problem that hope is searched by the mobile phone terminal access site.After question answering system is carried out participle to problem, utilize behind grammer and the part of speech information extraction keyword to search in the inverted index storehouse and comprise these crucial word problems, the information that finds out comprises that question and answer are to, the existing global frequencies information F of problem in knowledge base, simultaneously, each problem can have a keyword coverage rate C.Select the keyword coverage rate and be 60% problem as Candidate Set, the frequency collector unit obtains the positive frequency information f of this enquirement, and method is, gives positive integer between the 0-10 in proportion with these problems, coverage 60% be 1, coverage 100% be 10.The positive frequency information f is updated in the knowledge base by the frequency maintenance unit.Be the positive frequency information f of existing this enquirement of global frequencies information F+ of new global frequencies information F=, that is to say, need dynamically update user interactions behavior characterization parameter.
Sequencing unit is that index sorts to the problem information in the problem Candidate Set according to the product of new global frequencies F and coverage rate C, and the problem that ranks the first is returned to the user by the WAP website.
At this moment, if the user thinks that the problem of coupling is an erroneous matching, can click the button that reports an error on the WAP page, the next problem that system returns in the Candidate Set is automatically given the user.After simultaneously the frequency collector unit multiply by a penalty factor to the coverage of the aforementioned problem that is reported an error, be updated in the knowledge base by the frequency maintenance unit.The rest may be inferred, all do not find answer if Candidate Set all is a sky, and then the automatic recording user of system is putd question to pending tabulation.
Above content be in conjunction with concrete embodiment to further describing that the present invention did, can not assert that concrete enforcement of the present invention is confined to these explanations.For the general technical staff of the technical field of the invention, without departing from the inventive concept of the premise, can also make some simple deduction or replace, all should be considered as belonging to protection scope of the present invention.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010101662754A CN102236677A (en) | 2010-04-28 | 2010-04-28 | Question answering system-based information matching method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010101662754A CN102236677A (en) | 2010-04-28 | 2010-04-28 | Question answering system-based information matching method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102236677A true CN102236677A (en) | 2011-11-09 |
Family
ID=44887333
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010101662754A Pending CN102236677A (en) | 2010-04-28 | 2010-04-28 | Question answering system-based information matching method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102236677A (en) |
Cited By (31)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102831207A (en) * | 2012-08-06 | 2012-12-19 | 北京小米科技有限责任公司 | Computer terminal and information interaction method |
| CN103400054A (en) * | 2013-08-27 | 2013-11-20 | 哈尔滨工业大学 | Computer-assisted psychological consulting automatic question-answering robot system |
| WO2015007141A1 (en) * | 2013-07-16 | 2015-01-22 | International Business Machines Corporation | Correlating corpus/corpora value from answered questions |
| CN104572868A (en) * | 2014-12-18 | 2015-04-29 | 清华大学 | Method and device for information matching based on questioning and answering system |
| CN104572998A (en) * | 2015-01-07 | 2015-04-29 | 北京云知声信息技术有限公司 | Updating method and device of question answer sequencing model for automatic question answer system |
| CN104615755A (en) * | 2015-02-12 | 2015-05-13 | 北京航空航天大学 | Crowdsourcing-based novel question answering system |
| 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 |
| CN104951428A (en) * | 2014-03-26 | 2015-09-30 | 阿里巴巴集团控股有限公司 | User intention recognition method and device |
| CN105144205A (en) * | 2013-04-29 | 2015-12-09 | 西门子公司 | Device and method for answering a natural language question using a number of selected knowledge bases |
| CN105450781A (en) * | 2016-01-12 | 2016-03-30 | 李劲杰 | Interactive question and answer method, platform, client and server |
| CN105915490A (en) * | 2015-11-18 | 2016-08-31 | 乐视致新电子科技（天津）有限公司 | Security question interaction method and system |
| CN106055641A (en) * | 2016-05-31 | 2016-10-26 | 北京光年无限科技有限公司 | Human-computer interaction method and device oriented to intelligent robot |
| CN106095941A (en) * | 2016-06-14 | 2016-11-09 | 华南师范大学 | Recommend method and system based on the solution in big data knowledge storehouse |
| US9754207B2 (en) | 2014-07-28 | 2017-09-05 | International Business Machines Corporation | Corpus quality analysis |
| CN107180080A (en) * | 2017-04-28 | 2017-09-19 | 北京神州泰岳软件股份有限公司 | A kind of intelligent answer method and device of many interactive modes |
| CN108171347A (en) * | 2016-12-07 | 2018-06-15 | 北京京东尚科信息技术有限公司 | The implementation method and device of a kind of appointment business |
| CN108228637A (en) * | 2016-12-21 | 2018-06-29 | 中国电信股份有限公司 | Natural language client auto-answer method and system |
| CN108256056A (en) * | 2018-01-12 | 2018-07-06 | 广州杰赛科技股份有限公司 | Intelligent answer method and system |
| CN108920654A (en) * | 2018-06-29 | 2018-11-30 | 泰康保险集团股份有限公司 | A kind of matched method and apparatus of question and answer text semantic |
| CN108959257A (en) * | 2018-06-29 | 2018-12-07 | 北京百度网讯科技有限公司 | A kind of natural language analytic method, device, server and storage medium |
| CN109710749A (en) * | 2019-01-22 | 2019-05-03 | 深圳追一科技有限公司 | A kind of customer service auxiliary device and method |
| CN109783510A (en) * | 2019-01-22 | 2019-05-21 | 深圳追一科技有限公司 | Customer service auxiliary robot, method and system |
| CN110059171A (en) * | 2019-04-12 | 2019-07-26 | 中国工商银行股份有限公司 | Intelligent answer performance improvement method and system |
| CN110188205A (en) * | 2019-05-08 | 2019-08-30 | 三角兽（北京）科技有限公司 | A kind of update method and device of intelligent customer service system knowledge base |
| CN110837586A (en) * | 2018-08-15 | 2020-02-25 | 阿里巴巴集团控股有限公司 | Question-answer matching method, system, server and storage medium |
| CN111159363A (en) * | 2018-11-06 | 2020-05-15 | 航天信息股份有限公司 | Knowledge base-based question answer determination method and device |
| CN111382256A (en) * | 2020-03-20 | 2020-07-07 | 北京百度网讯科技有限公司 | Information recommendation method and device |
| CN111813911A (en) * | 2020-06-30 | 2020-10-23 | 神思电子技术股份有限公司 | A knowledge automatic collection and update system based on user supervision feedback and its working method |
| CN112231452A (en) * | 2020-10-12 | 2021-01-15 | 平安信托有限责任公司 | Question answering method, device, device and storage medium based on natural language processing |
| CN114548314A (en) * | 2022-03-02 | 2022-05-27 | 杭州网易再顾科技有限公司 | Text matching method and device, storage medium and electronic equipment |
| CN116414940A (en) * | 2021-12-28 | 2023-07-11 | 中国移动通信有限公司研究院 | Method, device and related equipment for determining standard problems |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101169797A (en) * | 2007-11-30 | 2008-04-30 | 朱廷劭 | Search result optimization method |
| CN101256573A (en) * | 2008-04-01 | 2008-09-03 | 天津大学 | Feedback Search Method Based on Content Correlation and Content Correlation Method |
| CN101320375A (en) * | 2008-07-04 | 2008-12-10 | 浙江大学 | Digital Book Search Method Based on User Click Behavior |
| CN101523338A (en) * | 2005-03-18 | 2009-09-02 | 搜索引擎科技有限责任公司 | Search engine that applies feedback from users to improve search results |
| CN101650731A (en) * | 2009-08-31 | 2010-02-17 | 浙江大学 | Method for generating suggested keywords of sponsored search advertisement based on user feedback |
-
2010
- 2010-04-28 CN CN2010101662754A patent/CN102236677A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101523338A (en) * | 2005-03-18 | 2009-09-02 | 搜索引擎科技有限责任公司 | Search engine that applies feedback from users to improve search results |
| CN101169797A (en) * | 2007-11-30 | 2008-04-30 | 朱廷劭 | Search result optimization method |
| CN101256573A (en) * | 2008-04-01 | 2008-09-03 | 天津大学 | Feedback Search Method Based on Content Correlation and Content Correlation Method |
| CN101320375A (en) * | 2008-07-04 | 2008-12-10 | 浙江大学 | Digital Book Search Method Based on User Click Behavior |
| CN101650731A (en) * | 2009-08-31 | 2010-02-17 | 浙江大学 | Method for generating suggested keywords of sponsored search advertisement based on user feedback |
Cited By (45)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102831207B (en) * | 2012-08-06 | 2015-01-28 | 小米科技有限责任公司 | Computer terminal and information interaction method |
| CN102831207A (en) * | 2012-08-06 | 2012-12-19 | 北京小米科技有限责任公司 | Computer terminal and information interaction method |
| CN105144205A (en) * | 2013-04-29 | 2015-12-09 | 西门子公司 | Device and method for answering a natural language question using a number of selected knowledge bases |
| CN105144205B (en) * | 2013-04-29 | 2018-05-08 | 西门子公司 | Apparatus and method for answering natural language questions using multiple selected knowledge bases |
| WO2015007141A1 (en) * | 2013-07-16 | 2015-01-22 | International Business Machines Corporation | Correlating corpus/corpora value from answered questions |
| US9275115B2 (en) | 2013-07-16 | 2016-03-01 | International Business Machines Corporation | Correlating corpus/corpora value from answered questions |
| CN103400054A (en) * | 2013-08-27 | 2013-11-20 | 哈尔滨工业大学 | Computer-assisted psychological consulting automatic question-answering robot system |
| CN104951428A (en) * | 2014-03-26 | 2015-09-30 | 阿里巴巴集团控股有限公司 | User intention recognition method and device |
| US9754207B2 (en) | 2014-07-28 | 2017-09-05 | International Business Machines Corporation | Corpus quality analysis |
| US10169706B2 (en) | 2014-07-28 | 2019-01-01 | International Business Machines Corporation | Corpus quality analysis |
| 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 |
| CN104572868A (en) * | 2014-12-18 | 2015-04-29 | 清华大学 | Method and device for information matching based on questioning and answering system |
| CN104572868B (en) * | 2014-12-18 | 2017-11-03 | 清华大学 | The method and apparatus of information matches based on question answering system |
| CN104572998A (en) * | 2015-01-07 | 2015-04-29 | 北京云知声信息技术有限公司 | Updating method and device of question answer sequencing model for automatic question answer system |
| CN104572998B (en) * | 2015-01-07 | 2017-09-01 | 北京云知声信息技术有限公司 | Question and answer order models update method and device for automatically request-answering system |
| CN104615755A (en) * | 2015-02-12 | 2015-05-13 | 北京航空航天大学 | Crowdsourcing-based novel question answering system |
| CN104615755B (en) * | 2015-02-12 | 2017-12-01 | 北京航空航天大学 | A kind of new question answering system based on mass-rent |
| CN105915490A (en) * | 2015-11-18 | 2016-08-31 | 乐视致新电子科技（天津）有限公司 | Security question interaction method and system |
| CN105450781A (en) * | 2016-01-12 | 2016-03-30 | 李劲杰 | Interactive question and answer method, platform, client and server |
| CN106055641A (en) * | 2016-05-31 | 2016-10-26 | 北京光年无限科技有限公司 | Human-computer interaction method and device oriented to intelligent robot |
| CN106055641B (en) * | 2016-05-31 | 2020-01-14 | 北京光年无限科技有限公司 | Intelligent robot-oriented man-machine interaction method and device |
| CN106095941B (en) * | 2016-06-14 | 2020-03-20 | 华南师范大学 | Big data knowledge base-based solution recommendation method and system |
| CN106095941A (en) * | 2016-06-14 | 2016-11-09 | 华南师范大学 | Recommend method and system based on the solution in big data knowledge storehouse |
| CN108171347A (en) * | 2016-12-07 | 2018-06-15 | 北京京东尚科信息技术有限公司 | The implementation method and device of a kind of appointment business |
| CN108228637B (en) * | 2016-12-21 | 2020-09-04 | 中国电信股份有限公司 | Automatic response method and system for natural language client |
| CN108228637A (en) * | 2016-12-21 | 2018-06-29 | 中国电信股份有限公司 | Natural language client auto-answer method and system |
| CN107180080A (en) * | 2017-04-28 | 2017-09-19 | 北京神州泰岳软件股份有限公司 | A kind of intelligent answer method and device of many interactive modes |
| CN108256056A (en) * | 2018-01-12 | 2018-07-06 | 广州杰赛科技股份有限公司 | Intelligent answer method and system |
| CN108959257A (en) * | 2018-06-29 | 2018-12-07 | 北京百度网讯科技有限公司 | A kind of natural language analytic method, device, server and storage medium |
| CN108920654A (en) * | 2018-06-29 | 2018-11-30 | 泰康保险集团股份有限公司 | A kind of matched method and apparatus of question and answer text semantic |
| CN108959257B (en) * | 2018-06-29 | 2022-11-22 | 北京百度网讯科技有限公司 | A natural language analysis method, device, server and storage medium |
| CN110837586B (en) * | 2018-08-15 | 2023-05-26 | 阿里巴巴集团控股有限公司 | Question-answer matching method, system, server and storage medium |
| CN110837586A (en) * | 2018-08-15 | 2020-02-25 | 阿里巴巴集团控股有限公司 | Question-answer matching method, system, server and storage medium |
| CN111159363A (en) * | 2018-11-06 | 2020-05-15 | 航天信息股份有限公司 | Knowledge base-based question answer determination method and device |
| CN109710749A (en) * | 2019-01-22 | 2019-05-03 | 深圳追一科技有限公司 | A kind of customer service auxiliary device and method |
| CN109783510A (en) * | 2019-01-22 | 2019-05-21 | 深圳追一科技有限公司 | Customer service auxiliary robot, method and system |
| CN110059171A (en) * | 2019-04-12 | 2019-07-26 | 中国工商银行股份有限公司 | Intelligent answer performance improvement method and system |
| CN110188205A (en) * | 2019-05-08 | 2019-08-30 | 三角兽（北京）科技有限公司 | A kind of update method and device of intelligent customer service system knowledge base |
| CN111382256A (en) * | 2020-03-20 | 2020-07-07 | 北京百度网讯科技有限公司 | Information recommendation method and device |
| CN111382256B (en) * | 2020-03-20 | 2024-04-09 | 北京百度网讯科技有限公司 | Information recommendation method and device |
| CN111813911A (en) * | 2020-06-30 | 2020-10-23 | 神思电子技术股份有限公司 | A knowledge automatic collection and update system based on user supervision feedback and its working method |
| CN112231452A (en) * | 2020-10-12 | 2021-01-15 | 平安信托有限责任公司 | Question answering method, device, device and storage medium based on natural language processing |
| CN116414940A (en) * | 2021-12-28 | 2023-07-11 | 中国移动通信有限公司研究院 | Method, device and related equipment for determining standard problems |
| CN114548314A (en) * | 2022-03-02 | 2022-05-27 | 杭州网易再顾科技有限公司 | Text matching method and device, storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102236677A (en) | Question answering system-based information matching method and system | |
| EP3819792A2 (en) | Method, apparatus, device, and storage medium for intention recommendation | |
| CN108304444B (en) | Information query method and device | |
| CN104166668B (en) | News commending system and method based on FOLFM models | |
| US8880559B2 (en) | Location activity search engine computer system | |
| CN103049440B (en) | A kind of recommendation process method of related article and disposal system | |
| CN105701216A (en) | Information pushing method and device | |
| CN102314443B (en) | The modification method of search engine and system | |
| CN110276010B (en) | Weight model training method and related device | |
| CN112084150B (en) | Model training, data retrieval method, device, equipment and storage medium | |
| CN105808590A (en) | Search engine realization method as well as search method and apparatus | |
| CN103605665A (en) | Keyword based evaluation expert intelligent search and recommendation method | |
| CN103577416A (en) | Query expansion method and system | |
| WO2012142553A2 (en) | Identifying query formulation suggestions for low-match queries | |
| CN103425687A (en) | Retrieval method and system based on queries | |
| CN105759983A (en) | System and method for inputting text into electronic devices | |
| CN105243149B (en) | A kind of semantic-based web query recommended method and system | |
| CN102722498A (en) | Search engine and implementation method thereof | |
| KR101508583B1 (en) | Semantic searching system and method for smart device | |
| US11249993B2 (en) | Answer facts from structured content | |
| KR102280494B1 (en) | Method for providing internet search service sorted by correlation based priority specialized in professional areas | |
| CN103198072A (en) | Method and device for mining and recommendation of popular search word | |
| CN107562966A (en) | The optimization system and method based on intelligence learning for web page interlinkage retrieval ordering | |
| CN103106234A (en) | Searching method and device of webpage content | |
| CN103955480A (en) | Method and equipment for determining target object information corresponding to user |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20111109 |