Movatterモバイル変換


[0]ホーム

URL:


CN103593412B - A kind of answer method and system based on tree structure problem - Google Patents

A kind of answer method and system based on tree structure problem
Download PDF

Info

Publication number
CN103593412B
CN103593412BCN201310506948.XACN201310506948ACN103593412BCN 103593412 BCN103593412 BCN 103593412BCN 201310506948 ACN201310506948 ACN 201310506948ACN 103593412 BCN103593412 BCN 103593412B
Authority
CN
China
Prior art keywords
client
node
node identification
query
tree structure
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201310506948.XA
Other languages
Chinese (zh)
Other versions
CN103593412A (en
Inventor
杨毅
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Jingdong Century Trading Co Ltd
Beijing Jingdong Shangke Information Technology Co Ltd
Original Assignee
Beijing Jingdong Century Trading Co Ltd
Beijing Jingdong Shangke Information Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Jingdong Century Trading Co Ltd, Beijing Jingdong Shangke Information Technology Co LtdfiledCriticalBeijing Jingdong Century Trading Co Ltd
Priority to CN201310506948.XApriorityCriticalpatent/CN103593412B/en
Publication of CN103593412ApublicationCriticalpatent/CN103593412A/en
Application grantedgrantedCritical
Publication of CN103593412BpublicationCriticalpatent/CN103593412B/en
Activelegal-statusCriticalCurrent
Anticipated expirationlegal-statusCritical

Links

Classifications

Landscapes

Abstract

The invention discloses a kind of answer method based on tree structure problem and system, the present invention is according to the relevance between all problems service attribute and the different stage in all problems under affiliated business, construct tree structure, each node storage problem of tree structure and corresponding response message, and node identification is set for each node in tree structure.The query of tree structure interior joint and node identification are showed into client, after the node identification that client is sent by internet is received, the problem of node of corresponding node identification is stored and corresponding response message are supplied to client.The present invention provides more accurate response message for the query of client, improves the user experience of client.

Description

A kind of answer method and system based on tree structure problem
Technical field
The present invention relates to e-commerce field, more particularly to a kind of answer method and system based on tree structure problem.
Background technology
Ecommerce mainly uses internet to be engaged in commercial affairs or activity, during ecommerce is carried out, ecommerceEnterprise needs to set online customer service in network side, provides the online solution of the related query of various commercial affairs for client by internetAnswer, commercial affairs correlation query is related to the consulting of the association attributes of commercial product, e-commerce venture's correlated activation consulting and after saleService consultation etc..With being continuously increased for client consulting amount, limited online customer service can not meet needs, in order toThe ever-increasing consulting amount of client is met, answering system is set in network side, the various business of client proposition are automatically repliedRelated query.
When the answering system set in network side automatically replies the various business correlation query of client proposition, first,In consulting amount, the related query of commercial affairs that up to ten thousand clients are proposed can be replied simultaneously;Secondly, in time, can be with whole day 24Hour replys the related query of commercial affairs that client is proposed.
At present, the related query of commercial affairs that answering system is generally proposed using matching way acknowledged client end, mainly using twoPlant implementation method:
The first implementation method uses the answering system response based on way of search
Answering system based on way of search is received after the query of client transmission, and main utilize searches for key technology pairAfter query matching, find corresponding response message and be sent to client.
Search for key technology and mainly utilize reverse indexing mode, reverse indexing is called inverted index, and this data structure is depositedStore up information and the mapping relations of information position certainly in one or more documents, that is, full-text search.When response systemSystem is received after the query of client transmission, query index is set up using query information as text, using reverse indexing modeThe problem of being fitted on correspondence, is then sent to client by the response message of the set correspondence problem.In this process, queryInformation can be multiple phrases, by combining method of calling, and the accuracy rate that Secondary Ordering Algorithm improves matching is aided with index.
Answering system based on way of search has the answering system based on full-text index engine lucene, based on full text ropeThe models such as the solr of secondary encapsulation can also be carried out on engine lucene.
Fig. 1 is the method flow diagram that the answering system of the prior art based on way of search realizes response, its specific stepSuddenly it is:
Step 101, the query for receiving by internet client input;
Step 102, to received client query use morphological analysis Language Processing, obtain after Language Processing doubtAsk;
Step 103, to after Language Processing query carry out syntactic analysis, obtain query information;
Step 104, using query information as text query index is set up, the database into answering system finds matchingMultiple problems and corresponding response message;
Because query information can be multiple, thus index matching the problem of be also include multiple problems and it is corresponding shouldAnswer information;
Step 105, quadratic dependence sequence is carried out to obtained multiple problems and corresponding response, obtain setting numberThe problem of degree of correlation highest and corresponding response message;
Step 106, corresponding response message the problem of the degree of correlation highest for setting number by internet is sent to visitorFamily end.
Second of implementation method uses the answering system response based on artificial intelligence markup language
Artificial intelligence markup language (AIML, Artificial Intelligence Markup Language) is a kind ofCreate the XML voices of natural language ageng.AIML objects are made up of theme, classification and indexing unit, constitute AIML pairsThe data of elephant can be format data or nonformatted data, and format data is made up of character, and which includes symbolResponse message can be encapsulated in the template of definition by data and the coherent element for constituting AIML, AIML, may part in templateIt is made up of to need to parse in response label and obtains response message.
Fig. 2 is the method flow diagram of the answering system response of the prior art based on artificial intelligence markup language, and it hasBody step is:
Step 201, the query for receiving by internet client input;
Step 202, use standardization to received client query, obtain query information;
Step 203, using query information to AIML KnowledgeBase-queries match, be encapsulated in the response in the template that AIML is definedInformation;
Step 204, the response message progress template processing to being encapsulated in the template that AIML is defined, parse mark thereinLabel, obtain response message;
Step 205, corresponding response message the problem of the degree of correlation highest for setting number by internet is sent to visitorFamily end.
The response of answering system no matter is realized using which kind of above-mentioned scheme, is all that answering system is first doubted to client propositionAsk that progress is analyzed to obtain after query information, then match and obtain corresponding response message.But, on the one hand, when query is using ChineseDuring statement, its semantics recognition is relatively difficult, and a same phrase is expressed under different pragmatics and linguistic context to be meant completely notWith, the meaning expressed by Chinese is recognized, it is necessary to which a word is divided into several words composition, because Chinese has veryPowerful ambiguousness, so existing analytical technology all can not be effectively by a word cutting.On the other hand, above two schemeIt is to carry out matching problem and corresponding response by the way of matching using inverted index and one by one when realizing the response of answering systemInformation, this matching way does not consider the correlation between problem, but directly matches query information with problem, instituteWith matching the problem of be possible to not be the problem of client needs, corresponding response message nor client need responseInformation, causes matching degree to reduce.
Therefore, when realizing the response of answering system, how accurate response message is provided for client, improves clientIt is a urgent problem to be solved using experience.
The content of the invention
In view of this, the embodiment of the present invention provides a kind of answer method based on tree structure problem, and this method can beClient proposes that problem provides accurate response message.
The embodiment of the present invention also provides a kind of answering system based on tree structure problem, and the system can be client instituteProposition problem provides accurate response message.
To reach above-mentioned purpose, what what the present invention was implemented be specifically realized in:
A kind of answer method based on tree structure problem, this method includes:
According to the relevance between all problems service attribute and the different stage in all problems under affiliated business, structureTree structure, each node storage problem of tree structure and corresponding response message are made, and is each section in tree structurePoint sets node identification;
The query of tree structure interior joint and node identification are showed into client;
Receive the node identification that client is sent by internet;
The problem of node of corresponding node identification is stored and corresponding response message are supplied to client.
This method also includes:
The problem of associated nodes for providing the node are stored is selected for client, and the associated nodes that client is selected are storedThe problem of and corresponding response message be supplied to client.
Before the node identification that client is sent by internet is received, this method also includes:
Determine whether the query that client is sent includes node identification, if it is, performing the section of corresponding node identificationThe step of the problem of point storage and corresponding response message are supplied to client;Obtained if not, carrying out syntactic analysis to the queryTo query information, the node identification of query information matches is determined, the problem of node of corresponding node identification is stored and correspondinglyResponse message be supplied to client.
Before the process per se, in addition to:
Business datum is arranged, business datum correlation analysis is carried out, and carry by modes such as correlation rule and contextsTake problem, offering question service attribute.
A kind of answering system based on tree structure problem, the system includes:Associate management module, data retrieval module andDisplay module is associated, wherein,
Associate management module, for according to the relevance between all problems service attribute and in all problems in affiliated industryDifferent stage under business, constructs tree structure, each node storage problem of tree structure and corresponding response message, and be treeEach node in shape structure sets node identification;
Data retrieval module, for receiving the node identification that client is sent by internet, is looked into from associate management moduleThe problem of node for asking corresponding node identification is stored and corresponding response message;
Display module is associated, for the query of tree structure interior joint and node identification to be showed into client, by inquiryThe problem of node of corresponding node identification is stored and corresponding response message.
Data retrieval module, the problem of associated nodes for also providing for the node are stored selects for client, from associationThe problem of associated nodes that management module inquiry obtains client selection are stored and corresponding response message;
Associate display module, be additionally operable to by inquire about obtain client selection associated nodes store the problem of and it is corresponding shouldAnswer information and show client.
The data retrieval module, is also used to determine whether that the query that can be sent from client gets node identification;If it is not, carrying out syntactic analysis to the query obtains query information, the node identification of query information matches is determined, from associated tubeThe problem of node of node identification corresponding to reason module polls is stored and corresponding response message.
Also include data preparation module, for being arranged to business datum, carry out business datum correlation analysis, and by closingThe mode such as connection rule and context extracts problem, offering question service attribute.
As can be seen from the above scheme, the present invention is according to the relevance between all problems service attribute and in all problemsDifferent stage under affiliated business, constructs tree structure, each node storage problem of tree structure and corresponding response letterBreath, and node identification is set for each node in tree structure.The query of tree structure interior joint and node identification are showedClient, after the node identification that client is sent by internet is received, the node of corresponding node identification is storedProblem and corresponding response message are supplied to client.Also include:The problem of associated nodes for further providing for the node are storedFor client selection, the problem of associated nodes that client is selected are stored and corresponding response message are supplied to client.MoreFurther, if node identification can not be got in the query received from client, grammer point can be carried out to the queryObtained after analysis after query information, determine the node identification of query information matches, by asking that the node of corresponding node identification is storedTopic and corresponding response message are supplied to client.Determined because the system and method that the present invention is provided is identified by exclusive nodeProblem and relevant issues, guiding client completes related consulting of all the problems in the affiliated business of query, it is to avoid prior artIn process and query information matches problem process are analyzed to the query of client, it is to avoid due to accuracy of analysis is not high andIt is not high and influence the problem of final response message accuracy is not high with accuracy, so as to be provided more for the query of clientAccurate response message, improves the user experience of client.
Brief description of the drawings
Fig. 1 is the method flow diagram that the answering system of the prior art based on way of search realizes response;
Fig. 2 is the method flow diagram of the answering system response of the prior art based on artificial intelligence markup language;
Fig. 3 is the answer method flow chart provided in an embodiment of the present invention based on tree structure problem;
Fig. 4 is the answering system structural representation provided in an embodiment of the present invention based on tree structure problem;
Fig. 5 is the structural representation of tree structure problem provided in an embodiment of the present invention.
Embodiment
For the objects, technical solutions and advantages of the present invention are more clearly understood, develop simultaneously embodiment referring to the drawings, rightThe present invention is described in further detail.
In order to provide accurate response message, it is necessary to avoid of the prior art to client for client institute statement into questionQuery analysis process and query information matches problem process, it is to avoid because accuracy of analysis is not high and matching accuracy is not highInfluence the problem of final response message accuracy is not high.Therefore, the present invention is according to the relevance between all problems service attributeAnd the different stage in all problems under affiliated business, construct tree structure, each node storage problem of tree structure andCorresponding response message, and node identification is set for each node in tree structure.By the query of tree structure interior joint and sectionPoint identification shows client, after the node identification that client is sent by internet is received, by corresponding node identificationNode is the problem of store and corresponding response message is supplied to client.Also include:Further provide for the association section of the nodeThe problem of point storage, the problem of storing the associated nodes that client is selected and corresponding response message were provided for client selectionTo client.
Further, if node identification can not be got in the query received, grammer can be carried out to the queryObtained after analysis after query information, determine the node identification of query information matches, the node of corresponding node identification is storedProblem and corresponding response message are supplied to client.
The present invention identifies determination problem and relevant issues by exclusive node, and guiding client is completed in the affiliated business of queryRelated consulting of all the problems, it is to avoid the semantic ambiguity sex chromosome mosaicism that can be caused during analysis query, so as to be carried by clientGo out query and accurate response message is provided, improve the user experience of client.
Fig. 3 is the answer method flow chart provided in an embodiment of the present invention based on tree structure problem, and it is concretely comprised the following steps:
Step 301, according to the relevance between all problems service attribute and all problems under affiliated business notSame level, constructs tree structure, each node storage problem of tree structure and corresponding response message, and in tree structureEach node set node identification;
In this step, according to the service attribute of problem, using correlation rule, sorting technique or artificial notation methods to industryTypical problem and relevant issues under business are divided, and construct tree structure;
Step 302, the query of tree structure interior joint and node identification showed into client;
Step 303, receive the node identification that client is sent by internet;
Step 304, the problem of the node of corresponding node identification is stored and corresponding response message is supplied to client.
After the step 304, this method also includes:
Step 305, provide for client selection the problem of the associated nodes of the node are stored, the association that client is selectedThe problem of node is stored and corresponding response message are supplied to client.
Before step 303, this method can also include:
Step 3031, determine client send query whether include node identification, if it is, perform step 303;Such asIt is really no, then perform step 3033;
Step 3032, to the query carry out syntactic analysis obtain query information;
The step needs to carry out the query syntactic analysis, and is scanned for using ranking index, obtains corresponding doubtInformation is asked, at this moment, it is necessary to the text comprising query is pre-set in database, corresponding query letter is obtained in the textBreath;
Step 3033, the node identification for determining query information matches, the problem of node of corresponding node identification is storedAnd corresponding response message is supplied to client.
Specifically, the whole process of step 301 is performed by data preparation module and associate management module, wherein,
Data preparation module, for being arranged to business datum, carries out business datum correlation analysis, and pass through correlation ruleProblem, offering question service attribute are extracted with the mode such as context;
Associate management module, for constructing tree structure, according to the relevance between all problems service attribute and in instituteThe problematic different stage under affiliated business, problem and corresponding response message are imported into tree structure respective nodes,Problem and corresponding response message can be verified when importing, determine problem with the presence or absence of repetition;Herein, can be withTo addition or deletion of node in the tree structure that is constructed, for filling into or deleting problematic by new problem.
Specifically, 302 described in Fig. 3 are completed to the process of step 3033 by data retrieval module, including:FirstJudge to receive the query that client is sent, it is determined whether node identification can be got from the query, if it is, will be correspondingThe problem of node of node identification is stored and corresponding response message are supplied to client;If it is not, then being carried out to the querySyntactic analysis obtains query information, determines the node identification of query information matches, and the node of corresponding node identification is storedThe problem of and corresponding response message be supplied to client.
In the present invention, also there is association display module, for being shown to tree structure, show:Tree-like knotThe problem of node for the node identification that structure problem and corresponding response message, correspondence client are sent is stored and corresponding response letterBreath, and the associated nodes of the node the problem of store and query matching that corresponding response message and client are sent sectionThe problem of node of point identification is stored and corresponding response message, and the associated nodes of the node the problem of store and it is corresponding shouldAnswer information.
Fig. 4 is the answering system structural representation provided in an embodiment of the present invention based on tree structure problem, as illustrated,Including:Associate management module, data retrieval module and association display module, wherein,
Associate management module, for according to the relevance between all problems service attribute and in all problems in affiliated industryDifferent stage under business, constructs tree structure, each node storage problem of tree structure and corresponding response message, and be treeEach node in shape structure sets node identification;
Data retrieval module, for receiving the node identification that client is sent by internet, is looked into from associate management moduleThe problem of node for asking corresponding node identification is stored and corresponding response message;
Display module is associated, for the query of tree structure interior joint and node identification to be showed into client, by inquiryThe problem of node of corresponding node identification is stored and corresponding response message.
Within the system, in addition to data preparation module, for being arranged to business datum, business datum correlation point is carried outAnalysis, and problem, offering question service attribute are extracted by modes such as correlation rule and contexts.
Within the system, data retrieval module, the problem of associated nodes for also providing for the node are stored supplies clientSelection, the problem of associated nodes for obtaining client selection from associate management module polls are stored and corresponding response message;
Associate display module, be additionally operable to by inquire about obtain client selection associated nodes store the problem of and it is corresponding shouldAnswer information and show client.
Within the system, the data retrieval module, is also used to determine whether that the query that can be sent from client is obtainedTo node identification;If it is not, carrying out syntactic analysis to the query obtains query information, the node mark of query information matches is determinedKnow, the problem of being stored from the node of the node identification corresponding to associate management module polls and corresponding response message.
A specific example is lifted the embodiment of the present invention is described in detail
Fig. 5 is the structural representation of tree structure problem provided in an embodiment of the present invention:When client is sent out by internetWhen the query sent can find corresponding node from tree structure, the node itself and its associated all nodes are just inquired aboutThe problem of storage and corresponding response message, can be by node itself and its associated all when showing clientThe problem of node is stored and corresponding response message are shown using tree structure.Client can again be found according to displaying and savePoint, End-Customer end can obtain the root of tree structure, just can be complete to each leaf node from the root of tree structureSolve some traffic issues.So by way of based on tree structure query, the consulting of guiding client finishing service very well,Good user experience is obtained using the user of client.As shown in figure 5, when query is determined to be the problem of wanting inquiry business1, then problem 1 and corresponding response are shown, and whether prompting relevant issues 3,4 and 5 and corresponding response show;When receivingQuery is determined to be the problem of wanting inquiry business 3, then showing problem 3 and corresponding response, and the related query 8 and 9 of promptingResponse whether show;By that analogy until completing the business all problems and corresponding response.So drawn by tree structureThe mode of leading allows client to carry out the selection of node query, it is to avoid client directly inputs query, it is to avoid Chinese ambiguity problem.
The such scheme provided from the present invention can be seen that the present invention by analyzing a large number of services search record, sum upRelation between main type of service and the related indicator problem of the business and these typical problems, tree-like knot is formed by problemStructure.Business consultation is carried out by tree structure problem guiding client, client is by clicking on mode, it is possible to it is determined that selectedNode identification, inquire about corresponding response in tree structure by node identification, it is to avoid client provides non-type doubtThe ambiguity problem during matching asked and caused.One preferably consulting experience is obtained by tree-like guide, so as to be more willing to selection certainlyDynamic answering system carries out business consultation, and online customer service pressure can be so reduced to greatest extent.
By experiment, in one month, business consultation is carried out using answering system of the present invention based on tree structure problemAmount can account for more than the 50% of total business consulting amount, improve efficiency and user experience.
Preferred embodiment is lifted above, the object, technical solutions and advantages of the present invention is further described, instituteIt should be understood that the foregoing is merely illustrative of the preferred embodiments of the present invention, it is not intended to limit the invention, it is all the present invention'sSpirit and principle within, any modifications, equivalent substitutions and improvements made etc., should be included in protection scope of the present invention itIt is interior.

Claims (6)

CN201310506948.XA2013-10-242013-10-24A kind of answer method and system based on tree structure problemActiveCN103593412B (en)

Priority Applications (1)

Application NumberPriority DateFiling DateTitle
CN201310506948.XACN103593412B (en)2013-10-242013-10-24A kind of answer method and system based on tree structure problem

Applications Claiming Priority (1)

Application NumberPriority DateFiling DateTitle
CN201310506948.XACN103593412B (en)2013-10-242013-10-24A kind of answer method and system based on tree structure problem

Publications (2)

Publication NumberPublication Date
CN103593412A CN103593412A (en)2014-02-19
CN103593412Btrue CN103593412B (en)2017-10-03

Family

ID=50083553

Family Applications (1)

Application NumberTitlePriority DateFiling Date
CN201310506948.XAActiveCN103593412B (en)2013-10-242013-10-24A kind of answer method and system based on tree structure problem

Country Status (1)

CountryLink
CN (1)CN103593412B (en)

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN104408639A (en)*2014-10-222015-03-11百度在线网络技术(北京)有限公司Multi-round conversation interaction method and system
KR102598273B1 (en)*2015-09-012023-11-06삼성전자주식회사Method of recommanding a reply message and device thereof
CN105354185B (en)*2015-10-302018-05-18上海智臻智能网络科技股份有限公司 Method and device for question answering system mining associated questions
CN108833595B (en)*2015-12-242021-06-29上海智臻智能网络科技股份有限公司 Computer-readable storage medium for online customer service
CN106528613B (en)*2016-05-262018-11-27中科鼎富(北京)科技发展有限公司Intelligent answer method and device
CN107547354A (en)*2016-06-272018-01-05南京中兴软件有限责任公司Business datum provides method and device
CN106991112B (en)*2016-11-072021-05-25创新先进技术有限公司Information query method and device
CN106599604B (en)*2017-01-122019-10-08上海博集医疗科技有限公司Medical information intelligent interactive system and method
CN107391551B (en)*2017-06-062020-04-14广东广业开元科技有限公司Web service data analysis method and system based on data mining
CN107992459A (en)*2017-10-122018-05-04如是科技(大连)有限公司Resume generation method and device
CN111476597B (en)*2020-03-192025-08-29深圳赛安特技术服务有限公司 A method for processing resource estimation results and related equipment

Citations (4)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN1851700A (en)*2005-11-012006-10-25华为技术有限公司Inquiry method and system, and inquiry switching device
CN101166212A (en)*2006-10-192008-04-23华为技术有限公司A method and system for realizing graphic display of interactive voice response system
CN101635010A (en)*2009-08-212010-01-27腾讯科技(深圳)有限公司Dialogue processing method and dialogue system in dialogue community
CN102375835A (en)*2010-08-172012-03-14腾讯科技(深圳)有限公司Information searching system and method

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN1851700A (en)*2005-11-012006-10-25华为技术有限公司Inquiry method and system, and inquiry switching device
CN101166212A (en)*2006-10-192008-04-23华为技术有限公司A method and system for realizing graphic display of interactive voice response system
CN101635010A (en)*2009-08-212010-01-27腾讯科技(深圳)有限公司Dialogue processing method and dialogue system in dialogue community
CN102375835A (en)*2010-08-172012-03-14腾讯科技(深圳)有限公司Information searching system and method

Also Published As

Publication numberPublication date
CN103593412A (en)2014-02-19

Similar Documents

PublicationPublication DateTitle
CN103593412B (en)A kind of answer method and system based on tree structure problem
CN111708869B (en) Method and device for man-machine dialogue processing
CN106156365B (en) A method and device for generating a knowledge graph
CN108804521B (en)Knowledge graph-based question-answering method and agricultural encyclopedia question-answering system
CN107220352B (en) Method and device for constructing review graph based on artificial intelligence
CN105608218B (en)The method for building up of intelligent answer knowledge base establishes device and establishes system
KR102288249B1 (en) Information processing method, terminal, and computer storage medium
US20140351228A1 (en)Dialog system, redundant message removal method and redundant message removal program
CN116628173B (en)Intelligent customer service information generation system and method based on keyword extraction
US9710829B1 (en)Methods, systems, and articles of manufacture for analyzing social media with trained intelligent systems to enhance direct marketing opportunities
CN109345282A (en)A kind of response method and equipment of business consultation
CN114238573A (en)Information pushing method and device based on text countermeasure sample
CN108984650B (en)Computer-readable recording medium and computer device
CN110175227A (en)A kind of dialogue auxiliary system based on form a team study and level reasoning
CN111062220B (en)End-to-end intention recognition system and method based on memory forgetting device
WO2019232893A1 (en)Method and device for text emotion analysis, computer apparatus and storage medium
CN106446045A (en)Method and system for building user portrait based on conversation interaction
CN110377745B (en)Information processing method, information retrieval device and server
CN117453895B (en)Intelligent customer service response method, device, equipment and readable storage medium
CN117421398A (en)Man-machine interaction method, device, equipment and storage medium
CN111046151A (en)Message processing method and device
CN112035623A (en)Intelligent question and answer method and device, electronic equipment and storage medium
CN118467845B (en) Method for constructing intelligent interactive service system, website intelligent interactive method and device
CN110209804B (en)Target corpus determining method and device, storage medium and electronic device
CN114372454B (en)Text information extraction method, model training method, device and storage medium

Legal Events

DateCodeTitleDescription
C06Publication
PB01Publication
SE01Entry into force of request for substantive examination
SE01Entry into force of request for substantive examination
GR01Patent grant
GR01Patent grant

[8]ページ先頭

©2009-2025 Movatter.jp