CN105630938A - Intelligent question-answering system - Google Patents
Intelligent question-answering system Download PDFInfo
- Publication number
- CN105630938A CN105630938A CN201510976272.XA CN201510976272A CN105630938A CN 105630938 A CN105630938 A CN 105630938A CN 201510976272 A CN201510976272 A CN 201510976272A CN 105630938 A CN105630938 A CN 105630938A
- Authority
- CN
- China
- Prior art keywords
- answer
- user
- question
- storehouse
- professional
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/90335—Query processing
- G06F16/90344—Query processing by using string matching techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/9032—Query formulation
- G06F16/90332—Natural language query formulation or dialogue systems
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides an intelligent question-answering system, which comprises more than one question input module, a question bank, a question matching module, a general answer bank, more than one professional answer bank and a question answering module, wherein the question input module is used for inputting a question which needs to be answered and setting a source identifier for the question, the source identifier comprises at least more than two types of large user information, each large user also comprises small users, the large users are mutually independent, and the small users among different large users are independent; and the question answering module finds a corresponding question answer in the professional answer bank or the general answer bank according to the matching result of the question matching module and the source identifier of the question and feeds a corresponding answer back to an answer requirement party according to the source identifier. Through the subdivision of users and the subdivision of answered questions, the intelligent question-answering system can give targeted accurate answers for different question asking parties.
Description
Technical field
The present invention relates to intelligent robot question and answer, natural language processing (NLP) field, limited language material is processed in the collection that principle is directed, goes to cover a session context that determine, little scope, promotes language identification degree, concrete, it is according to trade classification, gathers the data processing some industry, build industry knowledge base, realize the intelligent answer of an industry, be particularly suitable for intelligent customer service system.
Background technology
Along with social development, the progress of computer technology, yearning and the demand of artificial intelligence are strengthened by the mankind day by day, and the enterprise of some frontline technologies is proposed the chat robots of oneself one after another, ice as little in Microsoft, the little degree of Baidu, Fructus Mali pumilae SIRI etc. Make a general survey of panoramic chat robots, it has been found that these robots rely on big data technique, mass data is processed to form knowledge base, finds the answer of matching problem in knowledge base.
Experiencing these chat robots one by one to find, these chat robots amusement chat feature is powerful, has the popularity being packed in online social circle, and in contrast, for the every profession and trade customer service reducing artificial pressure with greater need for these intelligent answers, its using value is extremely low.
Robot such as puts question to " red how much? ", often answer " you say how much how much is exactly " this humorous answer, can win that popular one laughs at, customer service but can not be allowed for answering the practical problem of user. It is seen that, it has a problem in that the enquirement to robot does not indicate the object of problem, all trades and professions have the product of redness, in mass knowledge storehouse, and the answer that this problem can be corresponding a lot of, sell colored meeting and answer 5 yuan, the meeting selling clothes answers 500 yuan, and perhaps that sells computer can answer 5000 yuan, and program often selects to make mistakes the minimum answer of probability, for customer service, user, such answer is all not intended to.
And for example, puing question to " what material vaporizer is " to robot, test finds that existing robot all can not correctly answer, because this problem is more professional, the talent only refrigerator being very familiar with knows the answer. This illustrates, even if constructing the knowledge base comprising magnanimity question and answer, it is also difficult to more comprehensively cover all trades and professions.
In sum, the current ubiquity user of chat robots puts question to too wide, puts question to object indefinite, is difficult to build with the language material that artificial mark is enough the problem of the high knowledge base of coverage, causes that matching rate is low, answer accurate not, it is difficult to business is applied.
Summary of the invention
The present invention improves chat robots, adopt a set of new problem answers matching process, have more specific aim, set Intelligent Service background, solve to apply in the question and answer of some specific area at present, as the chat robots such as online specific merchant customer service are answered a question the low problem of accuracy.
For solving some technical problems existed in prior art, the present invention is a kind of Intelligent Answer System, including: a problem input module, problem input module is used for inputting the problem needing to answer, and problem is set its source identification, the large user's information comprised in described source identification is at least two or more, and large user includes little user further below, it is independent of each other between large user, is independent between the little user of same large user; More than one specialty problem base, is usually in specialty problem base and semantic identical problem is classified as same problem according to certain rule by the professional problem of specific industry or user, the source identification of an at least corresponding large user of professional problem base; One problem matching module, finds, according to the problem of problem input module input, the problem matched with it in problem base; More than one specialty answer storehouse, the answer of storage is usually the professional answer of specific industry or user, and identical problem is likely to there is different answers in different professional answer storehouses, the source identification of an at least corresponding large user in professional answer storehouse; One question response module, according to the source identification of the matching result of problem matching module and problem find in the answer storehouse of its correspondence corresponding problem answer and according to source identification by corresponding answer feedback to the answer side of needs.
As the further improvement of above-mentioned Intelligent Answer System, also including a common question storehouse and a general answer storehouse, the problem in common question storehouse is corresponding with the answer in general answer storehouse; Some common question of common question library storage, are usually the above large user of at least five and all can use, and are also, according to certain rule, semantic identical problem is classified as same problem; Described problem matching module search matching problem time specialty problem base than common question storehouse priority match.
Further improvement as above-mentioned Intelligent Answer System, also include a corpus labeling processing module, it extracts the problem in original language material and is labeled processing, and identical problem semantic after mark and corresponding answer are indexed corresponding relation, eventually forms a described problem base part.
As the further improvement of above-mentioned Intelligent Answer System, primal problem language material is carried out word segmentation processing by described corpus labeling processing module, extracts important word and generates keyword phrases, uses keyword phrases to replace primal problem; In order to completely cover relevant issues, simultaneously avoid again matching irrelevant problem, the practical situation according to problem, in the process in Construct question storehouse, manually the matching way of problem is labeled as be fully retained, keyword retains, orderly keyword retains; Wherein it is fully retained and refers to that directly the original language material of use, as problem, is not changed; Keyword reservation refers to not requiring that the set of keyword put in order replaces original language material; The reservation of orderly keyword refers to and replaces original language material with a tactic set of keyword. Accordingly, in problem matching process, different matching way marks is carried out different disposal by matching module: being fully retained, the match is successful to need problem base problem ability identical with the enquirement of user; Keyword retains, and the match is successful only to need all keywords that the enquirement of user comprises problem base problem to get final product; Orderly keyword retains, and needs all keywords that the enquirement of user comprises problem base problem, and is the keyword arranged in order, and the match is successful for ability.
As the further improvement of above-mentioned Intelligent Answer System, described source identification is including at least a kind of in webpage, computer client, mobile client, server end and inputs the large user of this problem and little Customs Assigned Number.
At least part of as the further improvement of above-mentioned Intelligent Answer System, described professional problem base and corresponding professional answer storehouse is automatically form further according to the certain rule set in system after related data information in this network address by accessing particular web site reading; The described professional problem base automatically formed and specialty answer storehouse have the specific identifier of himself large user, and its particular content can be modified by manual type by large user.
As the further improvement of above-mentioned Intelligent Answer System, described particular web site is set by large user, and it can carry out increasing or deleting as required.
As the further improvement of above-mentioned Intelligent Answer System, described access particular web site is that the time point set in its management system according to large user is automatically obtained by system, can feed back success or failure information after terminating.
Further improvement as above-mentioned Intelligent Answer System, described professional answer storehouse can specifically be subdivided into the professional answer storehouse of more than a layer less further below, its access is associated with source identification, so same professional problem itself probably due to source is different and corresponding answer in different more particular specialty answer storehouses.
As the further improvement of above-mentioned Intelligent Answer System, described answer feedback is to behind answer needs side, and the accuracy of corresponding answer can be labeled by answer needs side; Answer in described general answer storehouse and/or specialty answer storehouse can be modified as required by large user, wherein after large user's content modification by general answer storehouse, this partial content reforms into the professional answer storehouse part of this large user, and the problem corresponding with this answer also can become a part for the professional problem base of this large user.
The present invention has more specific aim, set up specialty problem base and specialty answer storehouse, so more can answer particular problem targetedly, specialty problem base can also be subdivided into less professional problem base, can constantly segment down for specialty problem, the accuracy of such question answering will improve constantly, and the level of intelligence of whole system can improve constantly.
Accompanying drawing explanation
Fig. 1 is one structural representation of the present invention;
Fig. 2 is one structural representation of the present invention;
Fig. 3 is a kind of simple process figure answered a question of the present invention.
Detailed description of the invention
Below in conjunction with accompanying drawing, the present invention is described in detail.
As shown in Figure 1, the present invention is a kind of Intelligent Answer System one embodiment, including: problem input module, concrete input terminal can be a lot of, this problem input module is used for inputting the problem needing to answer, and problem is set its source identification, the large user's information comprised in described source identification is at least two or more, large user includes little user further below, it is independent of each other between large user, being independent between the little user of same large user, little user can be further subdivided into less user further below, such as micro-user; There are two professional problem bases, specialty problem base are usually semantic identical problem is classified as same problem according to certain rule by the professional problem of specific industry or user, the source identification of an at least corresponding large user of professional problem base; There is a problem matching module, in problem base, find, according to the problem of problem input module input, the problem matched with it; Having two professional answer storehouses, the answer of storage is usually the professional answer of specific industry or user, and identical problem is likely to there is different answers in different professional answer storehouses, the source identification of an at least corresponding large user in professional answer storehouse; One question response module, according to the source identification of the matching result of problem matching module and problem find in the answer storehouse of its correspondence corresponding problem answer and according to source identification by corresponding answer feedback to the answer side of needs.
Large user and little user are a kind of identity, it is possible to representing by the particular number of bits of character, numeral, even one byte, it freely can set according to concrete practical application, as long as identity effect can be played. One large user can have numerous little user, and in this Intelligent Answer System, the user of highest level is large user, and the user of system level does not embody at this, also seldom sets forth. Such as a network selling platform itself just can become a large user in native system, and want on this platform buy product and by problem input module proposition counseling problem user be exactly little user, the problem that little user is proposed by different terminals all can embody the information such as little user, large user, input terminal in the present system. Little user can also be further subdivided into small user in fact, whether segment and determine according to applicable cases, this also for no other reason than that level and define, for instance be a network selling platform equally, itself be the large user of native system, a lot of little businessmans are also had below large user, little businessman itself is little user, and little businessman has small user further below, and these small users are only final specifically used person, it proposes the problem needing to solve, and is given and its answer by native system.
Such as Fig. 2 is Intelligent Answer System another kind embodiment schematic diagram of the present invention, including: problem input module, concrete input terminal can be a lot of, this problem input module is used for inputting the problem needing to answer, and problem is set its source identification, the large user's information comprised in described source identification is at least two or more, large user includes little user further below, it is independent of each other between large user, being independent between the little user of same large user, little user can be further subdivided into less user further below; Two professional problem bases, are usually in specialty problem base and semantic identical problem are classified as same problem according to certain rule by the professional problem of specific industry or user, the source identification of an at least corresponding large user of professional problem base; One problem matching module, finds, according to the problem of problem input module input, the problem matched with it in problem base; Two professional answer storehouses, the answer of storage is usually the professional answer of specific industry or user, and identical problem is likely to there is different answers in different professional answer storehouses, the source identification of an at least corresponding large user in professional answer storehouse; One question response module, according to the source identification of the matching result of problem matching module and problem find in the answer storehouse of its correspondence corresponding problem answer and according to source identification by corresponding answer feedback to the answer side of needs; Also including a common question storehouse and a general answer storehouse, the problem in common question storehouse is corresponding with the answer in general answer storehouse; Some common question of common question library storage, are usually the above large user of at least five and all can use, and are also, according to certain rule, semantic identical problem is classified as same problem; Described problem matching module search matching problem time specialty problem base than common question storehouse priority match.
Although Fig. 1, Fig. 2, Fig. 3 do not embody, actually the Intelligent Answer System of the present invention may also include a corpus labeling processing module, it extracts the problem in original language material and is labeled processing, identical problem semantic after mark and corresponding answer are indexed corresponding relation, eventually form specialty problem base or common question storehouse part or all; Primal problem language material is carried out word segmentation processing by corpus labeling processing module, extracts important word and generates keyword phrases, uses keyword phrases to replace primal problem; In order to completely cover relevant issues, simultaneously avoid again matching irrelevant problem, the practical situation according to problem, in the process in Construct question storehouse, manually the matching way of problem is labeled as be fully retained, keyword retains, orderly keyword retains; Wherein it is fully retained and refers to that directly the original language material of use, as problem, is not changed; Keyword reservation refers to not requiring that the set of keyword put in order replaces original language material; The reservation of orderly keyword refers to and replaces original language material with a tactic set of keyword. Accordingly, in problem matching process, different matching way marks is carried out different disposal by matching module: being fully retained, the match is successful to need problem base problem ability identical with the enquirement of user; Keyword retains, and the match is successful only to need all keywords that the enquirement of user comprises problem base problem to get final product; Orderly keyword retains, and needs all keywords that the enquirement of user comprises problem base problem, and is the keyword arranged in order, and the match is successful for ability.
Input terminal in Fig. 3 can be multiple, for instance a kind of in webpage, computer client, mobile client, server end and input the large user of this problem and little Customs Assigned Number, and concrete input terminal can be not limited to above-mentioned scope.
In Fig. 1 and Fig. 2, the mode of setting up in specialty answer storehouse has multiple, it can be specialty problem base and corresponding professional answer storehouse, at least part of is by accessing particular web site and read in this network address after related data information, automatically forming further according to the certain rule set in system; The professional problem base so automatically formed and specialty answer storehouse have the specific identifier of himself large user, and its particular content can be modified by manual type by large user, so makes answer more accurate.
The problem of problem input module input can pass through certain processing, as shown in Figure 3, different user inputs problem by different terminals, the source identification composition input problem string that different problems and user can be set by system, then just these input problems serially add in input rank, sequentially or according to certain rule extract an input problem string from input rank, matching problem is determined in source identification further according to input problem string, then it chooses corresponding answer storehouse, question response module selects coupling answer in corresponding general answer storehouse and specialty answer storehouse, the answer of coupling is exported to user.
Problem base in Fig. 2 is a common question storehouse and two professional problem base compositions, some common question of common question library storage in reality, is usually more than at least 5 industries and all can use; Although number Fig. 2 of specialty problem base only has two, lower tens of a lot of situations, individual simply basis up to a hundred in practical application, can be thousands of when many, the problem of its storage is usually the professional problem of specific industry or user, the problem that the form of expression is identical is likely to there is different matching problems in different professional problem bases, and then answer also differs.
In Fig. 1, Fig. 2 and Fig. 3, the mode of setting up in specialty answer storehouse has multiple, can be manually build storehouse, can also be automatically form further according to the certain rule set in system after related data information by accessing particular web site read in this network address, the speciality dictionary automatically formed can also have the specific identifier of himself large user, and this speciality dictionary entire content can be modified by manual type by large user. The storehouse automatically formed generally by large user to operate realization, can be automatically obtained by system by the mode of setting particular web site, network address can carry out increasing or deleting as required, and it can be automatically obtained by system according to the time point of setting in system that this setting accesses, for instance a period of time (such as 1 day or 2 days) goes to update a related content. Speciality dictionary can also segment further, it is possible to little user is divided into less user, and answer is more targeted, and these all have embodiment in source identification.
Fig. 1, general answer storehouse or the content in specialty answer storehouse in Fig. 2 and Fig. 3 can be arranged as required to into and by answer needs side, its accuracy can be labeled, these are labeled with being beneficial to system operator reference, the answer accuracy of raising system has positive role, simultaneously general answer storehouse or specialty answer storehouse in answer can be modified as required by large user, wherein after large user's content modification by general answer storehouse, this partial content can become the professional answer storehouse part of this large user, certainly this partial content can also become new general answer storehouse content because it has very strong versatility.
Download online original language material:
For the self-service customer service of agricultural bank, enter the self-service customer service page of agricultural bank. The left side of webpage is question and answer classification, is divided into " e-bank ", " bank card ", eight big classes such as " individual service ", click each big class, can "+" number launch, it is shown that the group below big class, the right of webpage is the problem that corresponding group comprises, and clicks problem, it is possible to display answer.
By web page tools such as Googles, find<divclass=" collapsed ">
<spanid="2">bank card</span>such webpage format, it is possible to analyze, before big class name, has<divclass="collapsed">
<spanid="2">such character string, by searching for this character string in webpage source code, it is possible to obtaining all of big class name, immediately following big class name, it is simply that little class name, form is such,<ul>
<li><ahref="/Kb/Category/37/2 pn=1&m=2 ">debit card
</a></li><li><ahref="/Kb/Category/38/2?pn=1&m=2">
The credit card</a></li>, namely the value of href is the URL of group, captures each group URL immediately following big class name, it is possible to obtain the URL of multiple problem.
Accessing group URL successively, obtain the problem that each group comprises, further according to said method, the URL of record problem, this URL further comprises the answer of problem. So, we just obtain all of problem and answer in whole self-service customer service column.
It is inner that these problems will be added to industry knowledge base (finance-bank-agricultural bank), before adding knowledge base to, the existing knowledge base of first retrieval, if problem exists, the then answer of replacement problem, ensure the ageing of question and answer with up-to-date answer, without the problem of coupling, then add question and answer to knowledge base as new knowledge.
After capturing online data, we get off the rule of this website acquisition question and answer and process recording, in order to next time, Automatic Program captured, and accumulated through after a while, it is possible to achieve regularly batch updating specialized knowledge base every time.
Branch trade data acquisition can be carried out during system during the concrete application present invention, for the industry that large user and cooperation businessman, enterprise are engaged in, by searching for network of relation question and answer, the channel such as customer service chat record, collect the substantial amounts of original language material of industry question and answer. From original language material, extract a number of question and answer as sample, form a limited FAQ. The quantity of sample can be 10,000 question and answer, it is also possible to is 100,000 question and answer, determines according to practical situation, and principle is that the more many effects of sample size are more good. Owing to substantial amounts of question and answer content is only limited to the online customer service of certain industry, therefore limited FAQ can effectively cover most key words that the sector customer service relates to.
In units of industry, labeled data builds specialized knowledge base, it is possible to reference to by being handled as follows:
(1) by multiple problems identical for semanteme, and a common answer is put together, forms a response unit. Substantial amounts of response unit set, constitutes an industry question and answer packet. After making a number of industry question and answer packet, response unit identical for problem in every profession and trade packet is extracted, forms a professional problem base and a professional answer storehouse corresponding thereto. Increase data-reusing, reduce artificial expense. Inter-trade common question, can be placed on common question storehouse, and answer is placed in general answer storehouse.
(2) to problem of different nature, " not needing amendment ", " must revise " to be marked. Some problem, for different enterprises, answer is different, such as " what cold-producing medium your refrigerator uses? " there are different answer in different enterprises, at this moment need in answer mark " must revise ", for " which cold-producing medium refrigerator can use? " such problem, then can mark " not needing amendment " in answer.
(3) problem of original language material is carried out word segmentation processing by programming. Such as " what if the refrigerator that the day before yesterday buys does not freeze? " after participle, become " what if the refrigerator that the day before yesterday buys does not freeze " eight words. Current participle technique and scheme company are a lot, as CRF, magnanimity, the Chinese Academy of Sciences ICTCLAS etc., pluses and minuses are different, do not enumerate in detail here.
(4) to the problem after participle, manually mark. According to practical situation, manually mark is divided into three classes, one is be fully retained, and directly replaces the problem in original language material by word segmentation result, as replace with " what if the refrigerator that the day before yesterday buys does not freeze " eight words in original language material problem " what if the refrigerator that the day before yesterday buys does not freeze? " Two is that keyword retains, and keyword important in artificial judgment word segmentation result replaces original language material with these keywords. As replace with " how refrigerator regulates temperature " four words in original language material problem " how the refrigerator that the day before yesterday buys regulates temperature? ", this set of keyword does not emphasize order, as " how regulating refrigerator temperature " also may be used; Three be ordered into keyword retain, it is impossible to change keyword order, as can only be " what if refrigerator does not freeze " replace original language material " what if the refrigerator that the day before yesterday buys does not freeze? "
According to default industry identification user's specialty problem, after referring to the enquirement accepting user, in specialty problem base, search the problem identical with user's question sentence, as found same problem, it is simply that a successful match. Specific to the present invention, it is simply that according to the industry that customer service pre-sets, in universal question-answer packet and corresponding professional industry question and answer packet, it is handled as follows:
(1) the question sentence participle to user, the set of participle, is called user's phrase.
(2) extracting the problem in packet and answer successively, problem has replaced to keyword phrases, and the keyword phrases type according to artificial mark carries out different disposal. If mark is " being fully retained ", then needing keyword phrases identical with user's phrase, the match is successful for ability; If mark be " keyword reservations ", as long as then contain all keywords of keyword phrases at user's phrase, get final product the match is successful; If mark is " orderly keyword reservation ", then needs user's phrase not only to comprise all keywords of keyword phrases, and the order of keyword can not change.
(3) by successful match to the answer of problem find in corresponding answer storehouse, question and answer is usually sets up certain index relative, so searching to get up to be easier, then by answer feedback to user.
These are only the preferably case study on implementation of the present invention, be not limited to the present invention, for a person skilled in the art, the present invention can have various modifications and variations. All within the spirit and principles in the present invention, any amendment of making, equivalent replacement, improvement etc., should be included within protection scope of the present invention.
Claims (10)
1. an Intelligent Answer System, it is characterised in that including:
More than one problem input module, problem input module is used for inputting the problem needing to answer, and problem is set its source identification, the large user's information comprised in described source identification is at least two or more, large user includes little user further below, it is independent of each other between large user, is independent between the little user of same large user;
More than one specialty problem base, is usually in specialty problem base and semantic identical problem is classified as same problem according to certain rule by the professional problem of specific industry or user, the source identification of an at least corresponding large user of professional problem base;
One problem matching module, finds, according to the problem of problem input module input, the problem matched with it in problem base;
More than one specialty answer storehouse, the answer of storage is usually the professional answer of specific industry or user, and identical problem is likely to there is different answers in different professional answer storehouses, the source identification of an at least corresponding large user in professional answer storehouse;
One question response module, according to the source identification of the matching result of problem matching module and problem find in the answer storehouse of its correspondence corresponding problem answer and according to source identification by corresponding answer feedback to the answer side of needs.
2. the Intelligent Answer System according to right 1, it is characterised in that:
Also including a common question storehouse and a general answer storehouse, the problem in common question storehouse is corresponding with the answer in general answer storehouse; Some common question of common question library storage, are usually the above large user of at least five and all can use, and are also, according to certain rule, semantic identical problem is classified as same problem; Described problem matching module search matching problem time specialty problem base than common question storehouse priority match.
3. the Intelligent Answer System according to right 1 or 2, it is characterised in that:
Also including a corpus labeling processing module, it extracts the problem in original language material and is labeled processing, and identical problem semantic after mark and corresponding answer are indexed corresponding relation, eventually forms a described problem base part.
4. the Intelligent Answer System according to right 3, it is characterised in that: primal problem language material is carried out word segmentation processing by described corpus labeling processing module, extracts important word and generates keyword phrases, uses keyword phrases to replace primal problem; In order to completely cover relevant issues, simultaneously avoid again matching irrelevant problem, the practical situation according to problem, in the process in Construct question storehouse, manually the matching way of problem is labeled as be fully retained, keyword retains, orderly keyword retains; Wherein it is fully retained and refers to that directly the original language material of use, as problem, is not changed; Keyword reservation refers to not requiring that the set of keyword put in order replaces original language material; The reservation of orderly keyword refers to and replaces original language material with a tactic set of keyword; Accordingly, in problem matching process, different matching way marks is carried out different disposal by matching module: being fully retained, the match is successful to need problem base problem ability identical with the enquirement of user; Keyword retains, and the match is successful only to need all keywords that the enquirement of user comprises problem base problem to get final product; Orderly keyword retains, and needs all keywords that the enquirement of user comprises problem base problem, and is the keyword arranged in order, and the match is successful for ability.
5. a kind of Intelligent Answer System according to right 1, it is characterised in that: described source identification is including at least a kind of in webpage, computer client, mobile client, server end and inputs the large user of this problem and little Customs Assigned Number.
6. the Intelligent Answer System according to right 1,2 or 5, it is characterised in that: described professional problem base and corresponding professional answer storehouse at least part of be automatically form further according to the certain rule set in system after related data information by accessing particular web site read in this network address; The described professional problem base automatically formed and specialty answer storehouse have the specific identifier of himself large user, and its particular content can be modified by manual type by large user.
7. the Intelligent Answer System according to right 6, it is characterised in that: described particular web site is set by large user, and it can carry out increasing or deleting as required.
8. the Intelligent Answer System according to right 6, it is characterised in that: described access particular web site is that the time point set in its management system according to large user is automatically obtained by system, can feed back success or failure information after terminating.
9. the Intelligent Answer System according to right 1, it is characterised in that: also have less micro-user, each one little user of micro-user's subordinate below described little user; Described professional answer storehouse can specifically be subdivided into the professional answer storehouse of more than a layer less further below, and its access is associated with source identification, so same professional problem itself probably due to source is different and corresponding answer in different more particular specialty answer storehouses.
10. the Intelligent Answer System according to right 1, it is characterised in that: described answer feedback is to behind answer needs side, and the accuracy of corresponding answer can be labeled by answer needs side; Answer in described general answer storehouse and/or specialty answer storehouse can be modified as required by large user, wherein after large user's content modification by general answer storehouse, this partial content reforms into the professional answer storehouse part of this large user, and the problem corresponding with this answer also can become a part for the professional problem base of this large user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510976272.XA CN105630938A (en) | 2015-12-23 | 2015-12-23 | Intelligent question-answering system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510976272.XA CN105630938A (en) | 2015-12-23 | 2015-12-23 | Intelligent question-answering system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105630938A true CN105630938A (en) | 2016-06-01 |
Family
ID=56045871
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510976272.XA Pending CN105630938A (en) | 2015-12-23 | 2015-12-23 | Intelligent question-answering system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105630938A (en) |
Cited By (28)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106021114A (en) * | 2016-06-02 | 2016-10-12 | 北京光年无限科技有限公司 | Automated testing method and system for intelligent robot |
| CN106202038A (en) * | 2016-06-29 | 2016-12-07 | 北京智能管家科技有限公司 | Synonym method for digging based on iteration and device |
| CN106326399A (en) * | 2016-08-22 | 2017-01-11 | 我骑我行（武汉）科技有限公司 | Data interaction system based on internet |
| CN106503242A (en) * | 2016-11-08 | 2017-03-15 | 上海智臻智能网络科技股份有限公司 | A kind of intelligent interactive method, device and server |
| CN106682221A (en) * | 2017-01-04 | 2017-05-17 | 上海智臻智能网络科技股份有限公司 | Response method and device for question and answer interaction and question and answer system |
| CN107301213A (en) * | 2017-06-09 | 2017-10-27 | 腾讯科技（深圳）有限公司 | Intelligent answer method and device |
| CN107368548A (en) * | 2017-06-30 | 2017-11-21 | 华中科技大学鄂州工业技术研究院 | Intelligent government affairs service interaction method and system |
| CN107423432A (en) * | 2017-08-03 | 2017-12-01 | 当家移动绿色互联网技术集团有限公司 | The method and system of professional problem and greeting problem are distinguished by robot |
| CN107862000A (en) * | 2017-10-22 | 2018-03-30 | 北京市农林科学院 | A kind of agricultural technology seeks advice from interactive method |
| CN108153875A (en) * | 2017-12-26 | 2018-06-12 | 广州蓝豹智能科技有限公司 | Language material processing method, device, intelligent sound box and storage medium |
| CN108460695A (en) * | 2018-03-28 | 2018-08-28 | 辛庆强 | A kind of agriculture and animal husbandry siphunculus reason systems and management method |
| CN108509476A (en) * | 2017-09-30 | 2018-09-07 | 平安科技（深圳）有限公司 | Problem associates method for pushing, electronic device and computer readable storage medium |
| CN108509609A (en) * | 2018-04-03 | 2018-09-07 | 广州幽联信息技术有限公司 | Intelligent human-machine interaction method, apparatus, computer equipment and storage medium |
| CN108595420A (en) * | 2018-04-13 | 2018-09-28 | 畅敬佩 | A kind of method and system of optimization human-computer interaction |
| CN108829777A (en) * | 2018-05-30 | 2018-11-16 | 出门问问信息科技有限公司 | A kind of the problem of chat robots, replies method and device |
| CN108920554A (en) * | 2018-06-20 | 2018-11-30 | 大国创新智能科技(东莞)有限公司 | Innovative approach and innovative education robot system based on big data and artificial intelligence |
| CN109271503A (en) * | 2018-11-06 | 2019-01-25 | 北京猎户星空科技有限公司 | Intelligent answer method, apparatus, equipment and storage medium |
| WO2019062001A1 (en) * | 2017-09-30 | 2019-04-04 | 平安科技（深圳）有限公司 | Intelligent robotic customer service method, electronic device and computer readable storage medium |
| 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 |
| CN110309378A (en) * | 2019-06-28 | 2019-10-08 | 深圳前海微众银行股份有限公司 | A kind of processing method that problem replies, apparatus and system |
| CN110472024A (en) * | 2019-07-11 | 2019-11-19 | 北京云迹科技有限公司 | For the configuration of the customized question and answer of robot, processing method and device, robot |
| CN110704589A (en) * | 2019-09-26 | 2020-01-17 | 浙江诺诺网络科技有限公司 | Data query method, device, equipment and computer readable storage medium |
| CN110990428A (en) * | 2019-12-17 | 2020-04-10 | 中国建设银行股份有限公司 | Problem solving method and device, electronic equipment and computer readable storage medium |
| CN112214654A (en) * | 2020-10-19 | 2021-01-12 | 厦门渊亭信息科技有限公司 | Universal intelligent question-answering automatic operation and maintenance system and method |
| TWI730536B (en) * | 2019-12-10 | 2021-06-11 | 中華電信股份有限公司 | A system for question recommendation and a method thereof |
| CN112966492A (en) * | 2021-02-09 | 2021-06-15 | 柳州智视科技有限公司 | Method for solving problem by using known knowledge |
| CN113302643A (en) * | 2018-10-10 | 2021-08-24 | 绍约公司 | System and method for multiple identification using smart contracts on blockchains |
| CN113449512A (en) * | 2020-03-25 | 2021-09-28 | 中国电信股份有限公司 | Information processing method, apparatus and computer readable storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001084376A2 (en) * | 2000-04-28 | 2001-11-08 | Global Information Research And Technologies Llc | System for answering natural language questions |
| CN102789496A (en) * | 2012-07-13 | 2012-11-21 | 携程计算机技术（上海）有限公司 | Method and system for implementing intelligent response |
| CN103902652A (en) * | 2014-02-27 | 2014-07-02 | 深圳市智搜信息技术有限公司 | Automatic question-answering system |
| CN104090907A (en) * | 2014-05-28 | 2014-10-08 | 北京百度网讯科技有限公司 | Method and device for providing information for user in specific application field |
| CN104850539A (en) * | 2015-05-28 | 2015-08-19 | 宁波薄言信息技术有限公司 | Natural language understanding method and travel question-answering system based on same |
-
2015
- 2015-12-23 CN CN201510976272.XA patent/CN105630938A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001084376A2 (en) * | 2000-04-28 | 2001-11-08 | Global Information Research And Technologies Llc | System for answering natural language questions |
| CN102789496A (en) * | 2012-07-13 | 2012-11-21 | 携程计算机技术（上海）有限公司 | Method and system for implementing intelligent response |
| CN103902652A (en) * | 2014-02-27 | 2014-07-02 | 深圳市智搜信息技术有限公司 | Automatic question-answering system |
| CN104090907A (en) * | 2014-05-28 | 2014-10-08 | 北京百度网讯科技有限公司 | Method and device for providing information for user in specific application field |
| CN104850539A (en) * | 2015-05-28 | 2015-08-19 | 宁波薄言信息技术有限公司 | Natural language understanding method and travel question-answering system based on same |
Cited By (33)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106021114B (en) * | 2016-06-02 | 2018-07-10 | 北京光年无限科技有限公司 | Towards the automated testing method and system of intelligent robot |
| CN106021114A (en) * | 2016-06-02 | 2016-10-12 | 北京光年无限科技有限公司 | Automated testing method and system for intelligent robot |
| CN106202038A (en) * | 2016-06-29 | 2016-12-07 | 北京智能管家科技有限公司 | Synonym method for digging based on iteration and device |
| CN106326399A (en) * | 2016-08-22 | 2017-01-11 | 我骑我行（武汉）科技有限公司 | Data interaction system based on internet |
| CN106503242A (en) * | 2016-11-08 | 2017-03-15 | 上海智臻智能网络科技股份有限公司 | A kind of intelligent interactive method, device and server |
| CN106503242B (en) * | 2016-11-08 | 2019-04-19 | 上海智臻智能网络科技股份有限公司 | A kind of intelligent interactive method, device and server |
| CN106682221A (en) * | 2017-01-04 | 2017-05-17 | 上海智臻智能网络科技股份有限公司 | Response method and device for question and answer interaction and question and answer system |
| CN107301213A (en) * | 2017-06-09 | 2017-10-27 | 腾讯科技（深圳）有限公司 | Intelligent answer method and device |
| WO2018224034A1 (en) * | 2017-06-09 | 2018-12-13 | 腾讯科技（深圳）有限公司 | Intelligent question answering method, server, terminal and storage medium |
| CN107368548A (en) * | 2017-06-30 | 2017-11-21 | 华中科技大学鄂州工业技术研究院 | Intelligent government affairs service interaction method and system |
| CN107423432A (en) * | 2017-08-03 | 2017-12-01 | 当家移动绿色互联网技术集团有限公司 | The method and system of professional problem and greeting problem are distinguished by robot |
| CN107423432B (en) * | 2017-08-03 | 2020-05-12 | 当家移动绿色互联网技术集团有限公司 | Method and system for distinguishing professional problems and small talk problems by robot |
| CN108509476A (en) * | 2017-09-30 | 2018-09-07 | 平安科技（深圳）有限公司 | Problem associates method for pushing, electronic device and computer readable storage medium |
| WO2019062001A1 (en) * | 2017-09-30 | 2019-04-04 | 平安科技（深圳）有限公司 | Intelligent robotic customer service method, electronic device and computer readable storage medium |
| CN107862000A (en) * | 2017-10-22 | 2018-03-30 | 北京市农林科学院 | A kind of agricultural technology seeks advice from interactive method |
| CN108153875A (en) * | 2017-12-26 | 2018-06-12 | 广州蓝豹智能科技有限公司 | Language material processing method, device, intelligent sound box and storage medium |
| CN108460695A (en) * | 2018-03-28 | 2018-08-28 | 辛庆强 | A kind of agriculture and animal husbandry siphunculus reason systems and management method |
| CN108509609A (en) * | 2018-04-03 | 2018-09-07 | 广州幽联信息技术有限公司 | Intelligent human-machine interaction method, apparatus, computer equipment and storage medium |
| CN108595420A (en) * | 2018-04-13 | 2018-09-28 | 畅敬佩 | A kind of method and system of optimization human-computer interaction |
| CN108829777A (en) * | 2018-05-30 | 2018-11-16 | 出门问问信息科技有限公司 | A kind of the problem of chat robots, replies method and device |
| CN108920554A (en) * | 2018-06-20 | 2018-11-30 | 大国创新智能科技(东莞)有限公司 | Innovative approach and innovative education robot system based on big data and artificial intelligence |
| CN108920554B (en) * | 2018-06-20 | 2020-12-22 | 大国创新智能科技(东莞)有限公司 | Creative method based on big data and artificial intelligence and creative education robot system |
| CN113302643A (en) * | 2018-10-10 | 2021-08-24 | 绍约公司 | System and method for multiple identification using smart contracts on blockchains |
| CN109271503A (en) * | 2018-11-06 | 2019-01-25 | 北京猎户星空科技有限公司 | Intelligent answer method, apparatus, equipment and storage medium |
| 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 |
| CN110309378A (en) * | 2019-06-28 | 2019-10-08 | 深圳前海微众银行股份有限公司 | A kind of processing method that problem replies, apparatus and system |
| CN110472024A (en) * | 2019-07-11 | 2019-11-19 | 北京云迹科技有限公司 | For the configuration of the customized question and answer of robot, processing method and device, robot |
| CN110704589A (en) * | 2019-09-26 | 2020-01-17 | 浙江诺诺网络科技有限公司 | Data query method, device, equipment and computer readable storage medium |
| TWI730536B (en) * | 2019-12-10 | 2021-06-11 | 中華電信股份有限公司 | A system for question recommendation and a method thereof |
| CN110990428A (en) * | 2019-12-17 | 2020-04-10 | 中国建设银行股份有限公司 | Problem solving method and device, electronic equipment and computer readable storage medium |
| CN113449512A (en) * | 2020-03-25 | 2021-09-28 | 中国电信股份有限公司 | Information processing method, apparatus and computer readable storage medium |
| CN112214654A (en) * | 2020-10-19 | 2021-01-12 | 厦门渊亭信息科技有限公司 | Universal intelligent question-answering automatic operation and maintenance system and method |
| CN112966492A (en) * | 2021-02-09 | 2021-06-15 | 柳州智视科技有限公司 | Method for solving problem by using known knowledge |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105630938A (en) | Intelligent question-answering system | |
| AU2015277425B2 (en) | Learning and using contextual content retrieval rules for query disambiguation | |
| US20210279232A1 (en) | Chatbot Search System, Chatbot Search Method, and Program | |
| US20110153595A1 (en) | System And Method For Identifying Topics For Short Text Communications | |
| CN105045847B (en) | A kind of method that Chinese institutional units title is extracted from text message | |
| CN103678362A (en) | Search method and search system | |
| CN112579729A (en) | Training method and device for document quality evaluation model, electronic equipment and medium | |
| CN109597934B (en) | Method and device for determining click recommendation words, storage medium and electronic equipment | |
| CN107515866A (en) | A kind of data manipulation method, device and system | |
| CN108255963A (en) | A kind of control method and device of the News Retrieval based on internet | |
| CN101763424A (en) | Method for determining characteristic words and searching according to file content | |
| CN105956168B (en) | A kind of information retrieval method and information indexing device | |
| CN107705849A (en) | Remote medical consultation with specialists opinion integration method and device | |
| CN110717008B (en) | Search result ordering method and related device based on semantic recognition | |
| US20210165835A1 (en) | Computer driven question identification and understanding within a commerical tender document for automated bid processing for rapid bid submission and win rate enhancement | |
| CN115760495A (en) | Method and device for realizing automatic labeling of legal cases | |
| CN106612336A (en) | Picture preloading method and picture preloading device | |
| CN109062986A (en) | A kind of classification processing method and device of label | |
| CN113590792A (en) | User problem processing method and device and server | |
| CN115114441A (en) | Query graph construction method and device, storage medium and electronic equipment | |
| CN105224596A (en) | A kind of method of visit data and device | |
| CN112749316A (en) | Translation quality determination method and device, storage medium and processor | |
| CN110825930A (en) | Method for automatically identifying correct answers in community question-answering forum based on artificial intelligence | |
| CN107423353A (en) | A kind of inquiry system based on first letter of pinyin | |
| CN113254622B (en) | Knowledge point query method, knowledge point query device and knowledge point query server |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160601 |