CN108228559A - A kind of human-computer interaction realization method and system for customer service - Google Patents
A kind of human-computer interaction realization method and system for customer service Download PDFInfo
- Publication number
- CN108228559A CN108228559A CN201611197153.5A CN201611197153A CN108228559A CN 108228559 A CN108228559 A CN 108228559A CN 201611197153 A CN201611197153 A CN 201611197153A CN 108228559 A CN108228559 A CN 108228559A
- Authority
- CN
- China
- Prior art keywords
- operation system
- information
- recognition result
- user
- chat content
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0613—Third-party assisted
Abstract
The embodiment of the invention discloses a kind of human-computer interaction realization method and systems for customer service, are related to intelligent robot technology field, can expand the application range of chat robots.The present invention includes:The chat content that user inputs user equipment is obtained, and the recognition result of user view is obtained according to the chat content;It determines the operation system corresponding to the recognition result, and accesses at least one operation system;According to the recognition result and the business information extracted from the operation system accessed, the operational order of the corresponding chat content of generation;According to the operational order generated, at least one operation system is called, and perform the business event of at least one operation system.The present invention is suitable for the application range of extension chat robots.
Description
Technical field
The present invention relates to intelligent robot technology field more particularly to a kind of human-computer interaction realization sides for customer service
Method and system.
Background technology
Chat robots are as a kind of chat agency, by running the computer program of chat robots, with the mankind with audio
Or the forms such as text engage in the dialogue.At present, chat robots are primarily used to customer service question and answer, Personal Assistant, phonetic search tool
The fields of grade as a kind of new mode of human-computer interaction, improve user and use the Experience Degree in line service to a certain extent.But
It is that chat robots really are able to bring the aspect of income for operator, it is main still by as a kind of reduction manual service
The tool of cost uses, and especially in customer service question and answer, by automatically replying some general answers, is proposed with filtration fraction user
Query, so as to reduce the cost of labor of customer service.
It is automatic to answer carrying online for customer by preset " problem base " in background server with the artificial example of customer service machine
It asks the workload so as to reduce contact staff, reduces human cost therein.Business event is based on the chatting machine of " problem base "
Device people mainly still applies the phonetic search device etc. in customer service question and answer, " release both hands " formula, such as:It is used for
Search engine carries out the retrieval of commodity.To reach the business event of the existing businesses systems such as current electric business platform, on-line finance,
Such as:It places an order, the business event such as order inquiries, logistics enquiring, then needs that user transfers manual service or user switchs to manually
Operation input etc. traditional mode of operation, to complete existing on-line operation behavior, limit chat robots apply model
It encloses.
Invention content
The embodiment of the present invention provides a kind of human-computer interaction realization method and system for customer service, can expand merely
The application range of its robot.
In order to achieve the above objectives, the embodiment of the present invention adopts the following technical scheme that:
In a first aspect, the method that the embodiment of the present invention provides, including:
The chat content that user inputs user equipment is obtained, and the identification knot of user view is obtained according to the chat content
Fruit;
It determines the operation system corresponding to the recognition result, and accesses at least one operation system;
According to the recognition result and the business information extracted from the operation system accessed, in the corresponding chat of generation
The operational order of appearance, wherein, the operational order generated, for calling at least one operation system.
With reference to first aspect, it in the first possible realization method of first aspect, further includes:
According to the operational order generated, at least one operation system is called, and perform at least one business
The business event of system.
With reference to first aspect, it is described according to the chat content in second of possible realization method of first aspect
The recognition result of user view is obtained, including:
Specific field is obtained from the character included by the chat content, and identifies the letter that the specific field includes
Breath, the specific field are the character field comprising specify information;
According to the specific field that identification obtains, the semanteme and structure of each character field in the chat content are determined;
According to the semanteme and structure of each character field in the identified chat content, described chat is determined by disaggregated model
The corresponding intention classification of its content, and obtain the recognition result.
With reference to first aspect, it in the third possible realization method of first aspect, further includes:
The chat content of user equipment is currently inputted according to user, obtains the recognition result that active user is intended to;
According to the recognition result that the active user is intended to, the operational order is updated.
The possible realization method of with reference to first aspect the first, in the 4th kind of possible realization method, the basis
The recognition result that the active user is intended to, updates the operational order, including:
According to the recognition result that the active user is intended to, switch the operation system of current accessed, and from the industry after switching
The business information extracted in business system;And it according to the business information extracted in the operation system after switching, updates the operation and refers to
It enables;
Alternatively, according to the recognition result that the active user is intended to, from the operation system of current accessed, behaviour is redefined
It instructs.
With reference to first aspect, it in the 5th kind of possible realization method of first aspect, further includes:
Answer model is extracted from the knowledge base, and is screened and obtained from the business information according to the answer model
Information to be feedback;
Response message is generated according to the recognition result and the information to be feedback, the response message generated is carried out
Screening obtains answer message, and sends answer message to the user equipment.
The 5th kind of possible realization method with reference to first aspect, in the 6th kind of possible realization method, the basis
The answer model screens from the business information and obtains information to be feedback, including:
The answer model of corresponding online transaction platform is extracted from the knowledge base, and is read from the online transaction platform
Business information, wherein, read business information includes:For completing merchandise display, shopping guide, product information consulting, buying and determine
Plan support, performance comparison, order generation, payment flow, logistic track, user feedback and follow-up customer service business information;
According to the answer model of the corresponding online transaction platform, the chat content of user equipment is inputted for user, from
Screening obtains information to be feedback in the business information, wherein, the content in the information to be feedback, and from the chat
Specific field is obtained in character included by content, there is the part mutually met.
The 5th kind of possible realization method with reference to first aspect, in the 7th kind of possible realization method, further includes:
According to the recognition result, reply data is read from the knowledge base, and looked into the reply data read
Inquiry obtains response message, and the type of the reply data includes at least:Document data, audio data and video data, the finger
Fixed response message includes:Text message, audio-frequency information or video information;
The obtained response message of inquiry is ranked up according to preset ordering rule, and is obtained according to ranking results described
Answer message.
The 7th kind of possible realization method with reference to first aspect, in the 8th kind of possible realization method, further includes:
To the business information extracted from each operation system, carried out according to each service node of the operation system in institute source
Cluster obtains cluster set, and stores into the knowledge base;
After corresponding at least one operation system is accessed, the cluster set of corresponding at least one operation system is established
The access interface of conjunction, and pass through the access interface, the business information of each service node in the operation system is read, and make
For the response message.
The 7th kind of possible realization method with reference to first aspect, in the 9th kind of possible realization method, it is described according to
The response message that preset ordering rule obtains inquiry is ranked up, including:
According to response message and the correlation and similitude identified between obtained specific field, information is calculated
Distance;
The response message obtained according to described information distance to inquiry is ranked up.
Second aspect, the system that the embodiment of the present invention provides, including:
The intention Analysis server, for obtaining the chat content that user inputs user equipment, and according to the chat
The recognition result of content obtaining user view;And determine the operation system corresponding to the recognition result, and described in access at least
One operation system;
The operation system, for sending business information to the intention Analysis server;
The intention Analysis server is additionally operable to according to the recognition result and the industry extracted from the operation system accessed
Be engaged in information, the operational order of the corresponding chat content of generation, wherein, the operational order generated, for call it is described at least
One operation system;And according to the operational order generated, call at least one operation system;
The operation system is additionally operable to the operational order according to the intention Analysis server generation, performs business event.
With reference to second aspect, in the first possible realization method of second aspect, the intention Analysis server has
Body is used to obtain specific field from the character included by the chat content, and identify the information that the specific field includes,
The specific field is the character field comprising specify information;And the specific field obtained according to identification, determine the chat content
In each character field semanteme and structure;Further according to the semanteme and structure of each character field in the identified chat content, pass through
Disaggregated model determines the corresponding intention classification of the chat content, and obtains the recognition result.
With reference to second aspect, in second of possible realization method of second aspect, the intention Analysis server is gone back
For currently inputting the chat content of user equipment according to user, the recognition result that active user is intended to is obtained;And according to described
The recognition result that active user is intended to, updates the operational order.
With reference to the first possible realization method of second aspect, in the third possible realization method, further include:Institute
Intention Analysis server is stated, specifically for the recognition result being intended to according to the active user, switches the business system of current accessed
System, and the business information extracted from the operation system after switching;And believed according to the business extracted in the operation system after switching
Breath, updates the operational order;Alternatively, according to the recognition result that the active user is intended to, from the operation system of current accessed
In, redefine operational order.
With reference to second aspect, in the 4th kind of possible realization method of second aspect, the intention Analysis server is gone back
For extracting answer model from the knowledge base, and screened from the business information according to the answer model and obtain treating instead
The information of feedback;And response message is generated according to the recognition result and the information to be feedback, to the response message generated
It is screened to obtain answer message, and answer message is sent to the user equipment.
With reference to the 4th kind of possible realization method of second aspect, in the 5th kind of possible realization method, further include:Institute
Intention Analysis server is stated, specifically for extracting the answer model of corresponding online transaction platform from the knowledge base, and from institute
It states online transaction platform and reads business information, wherein, read business information includes:For completing merchandise display, shopping guide, production
Product information consultation, purchase decision support, performance comparison, order generation, payment flow, logistic track, user feedback and follow-up visitor
The business information of clothes;And according to the answer model of the corresponding online transaction platform, the chat of user equipment is inputted for user
Content, from the business information screening obtain information to be feedback, wherein, the content in the information to be feedback, with from
Specific field is obtained in character included by the chat content, there is the part mutually met.
With reference to the 4th kind of possible realization method of second aspect, in the 6th kind of possible realization method, further include:Institute
Intention Analysis server is stated, is additionally operable to according to the recognition result, reply data is read from the knowledge base, and reading
Reply data in inquiry obtain response message, the type of the reply data includes at least:It document data, audio data and regards
Frequency evidence, the response message specified include:Text message, audio-frequency information or video information;And it is advised according to preset sequence
The response message then obtained to inquiry is ranked up, and obtain the answer message according to ranking results.
Human-computer interaction realization method and system provided in an embodiment of the present invention for customer service, can be by by chat
Hold identified user view, determine corresponding operation system, and the enquirement of user is converted into the identifiable operation of system and is referred to
It enables, and calls each operation system to reach the business event that can be provided in current operation system.Relative to base in the prior art
In the response mode of preset " problem base ", the recognition result the present invention is based on user view carries out response, and recalls symbol automatically
The business of family current demand is shared, and task, which is reached, to be realized to the subsequent business event of each operation system by chat robots.
The responsibility of chat robots is not only increased, but also capax negotii is provided, expands application range.
Description of the drawings
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to needed in the embodiment
Attached drawing is briefly described, it should be apparent that, the accompanying drawings in the following description is only some embodiments of the present invention, for ability
For the those of ordinary skill of domain, without creative efforts, it can also be obtained according to these attached drawings other attached
Figure.
Fig. 1 is the configuration diagram that system is realized in the human-computer interaction provided in an embodiment of the present invention for customer service;
Fig. 2 a are that the human-computer interaction provided in an embodiment of the present invention for customer service realizes that the flow of the method for system is shown
It is intended to;
Fig. 2 b are a kind of schematic diagram of specific example provided in an embodiment of the present invention;
Fig. 3, Fig. 4, Fig. 5 are that the human-computer interaction provided in an embodiment of the present invention for customer service realizes that the structure of system is shown
It is intended to.
Specific embodiment
For those skilled in the art is made to more fully understand technical scheme of the present invention, below in conjunction with the accompanying drawings and specific embodiment party
Formula is described in further detail the present invention.Embodiments of the present invention are described in more detail below, the embodiment is shown
Example is shown in the drawings, wherein same or similar label represents same or similar element or with identical or class from beginning to end
Like the element of function.It is exemplary below with reference to the embodiment of attached drawing description, is only used for explaining the present invention, and cannot
It is construed to limitation of the present invention.Those skilled in the art of the present technique are appreciated that unless expressly stated, odd number shape used herein
Formula " one ", "one", " described " and "the" may also comprise plural form.It is to be further understood that the specification of the present invention
The middle wording " comprising " used refers to there are the feature, integer, step, operation, element and/or component, but it is not excluded that
Other one or more features of presence or addition, integer, step, operation, element, component and/or their group.It should be understood that
When we claim element to be " connected " or during " coupled " to another element, it can be directly connected or coupled to other elements or
There may also be intermediary elements.In addition, " connection " used herein or " coupling " can include wireless connection or coupling.Here make
Wording "and/or" includes any cell of one or more associated list items and all combines.The art
Technical staff is appreciated that unless otherwise defined all terms used herein have (including technical terms and scientific terms)
The meaning identical with the general understanding of the those of ordinary skill in fields of the present invention.It should also be understood that such as general
Those terms defined in dictionary, which should be understood that, to be had a meaning that is consistent with the meaning in the context of the prior art, and
Unless being defined as here, will not be explained with the meaning of idealization or too formal.
The embodiment of the present invention can be specifically realized in a kind of system as shown in Figure 1, wherein:
It can be the server apparatus being individually made to be intended to Analysis Service implement body, such as:Rack, blade, it is tower or
The server apparatus of person's cabinet-type can also use work station, mainframe computer etc. to have stronger computing capability hardware device;
The server cluster that can be made of multiple server apparatus.It is intended to Analysis server for performing what the present embodiment was provided
Method flow, such as:Receive the text envelopes such as chat messages, short message that user equipment is sent by wireless network or internet
Breath or voice messaging, and send response message to user equipment.Multiple operation systems can be docked by being intended to Analysis server, from
Corresponding business information is extracted in operation system and generates specific response message.
Operation system can be specifically to be used for the operation systems such as online transaction business, financial business or logistics business, than
Such as it can include but is not limited to:Online shopping platform, order processing system, Order Query System, commodity parameter library, commodity price
Library, commodity information database, advertising campaign library, after-sale service system, inventory system, shopping cart management system, membership management system etc..
User equipment can specifically make an independent system or be integrated in a variety of different medium data playing systems in fact
In, such as smart mobile phone, tablet computer (Tablet Personal Computer), laptop computer (Laptop
Computer), personal digital assistant (personal digital assistant, abbreviation PDA) or wearable device
(Wearable Device) etc..Can chat interface be shown by the application program or APP of installation on user equipment, chatted
User equipment is specifically shown in its interface to letters such as chat messages, short message, the speech messages for being intended to Analysis server transmission
Breath, and show and be intended to the response message that Analysis server returns, so as to form user equipment and be intended between Analysis server
" user conversation ", and show the operation user of user equipment.
It is common, it is intended that Analysis server and each operation system can Interworking Data library system, Database Systems tool
Body can be individually made, management for data, the server apparatus of storage or by multiple server apparatus groups
Into server cluster.Corresponding Analysis server and each operation system of being intended to is run on the hardware device of Database Systems
Database, for managing and storing the data for being intended to Analysis server and each operation system.Common net specifically may be used
Shape database (Network Database), relational database (Relational Database), tree shaped data library
(Hierarchical Database), object-oriented database (Object-oriented Database) and a new generation are big
Data system framework.
The embodiment of the present invention provides a kind of human-computer interaction implementation method for customer service, as shown in Fig. 2 a, 2b, packet
It includes:
S1, the chat content that user inputs user equipment is obtained, and the knowledge of user view is obtained according to the chat content
Other result.
In the present embodiment, chat content by input equipments such as touch screen, keyboard, microphones can be inputted and used by user
Family equipment, the chat content inputted can be the data of the information formats such as text message, voice messaging and image information.Its
In, for the specifying information type of the chat content inputted, corresponding analysis and identification mode may be used, such as:If institute is defeated
The chat content entered is voice messaging, then can identify specific chat content by existing speech recognition technology, and according to
The chat content obtains the recognition result of user view;It, can be by existing if the chat content inputted is image information
Some image analysis technologies are distinguished and identify the specific object in image or have the image-region of specified meaning, so as to
It identifies specific chat content, and the recognition result of user view is obtained according to the chat content;If in the chat inputted
Hold for text message, then specific chat content can be identified by character recognition technology, and obtain according to the chat content
The recognition result of user view, such as:The concrete mode of the recognition result of user view, packet are obtained according to the chat content
It includes:
Specific field is obtained from the character included by the chat content, and identifies the letter that the specific field includes
Breath, wherein, the specific field is the character field comprising specify information, can be managed by the specific field extracted in chat content
It solves as representing the character of the intention of user.Such as:Natural language processing (Natural is carried out to the chat content
Language Processing), such as:The understanding of word input by user may be used " bag of words " model, according to
Part of speech, tense etc. are by the text of further label segmentation input;And language ambience information can be modeled further by word2vec, be led to
It crosses probabilistic language model and weight is assigned to vocabulary;The effect of natural language processing is further promoted based on depth neuroid again,
The professional dictionary for identifying character included in the chat content is obtained, such as:Various in electric business field include
The professional dictionary (such as brand dictionary, product dictionary, the type that are used to represent the character of the intention of user in each specific business scope
Number dictionary etc.).By these professional dictionaries, the identification for the information that the specific field includes.
And according to obtained specific field is identified, determine the semanteme and structure of each character field in the chat content.Root again
According to the semanteme and structure of each character field in the identified chat content, determine that the chat content corresponds to by disaggregated model
Intention classification, and obtain the recognition result.Such as:It comforms and rationality and coherent session is formed in the input of multi-user, and
The identification of behavior pattern of user and the evaluation and test of user view are carried out in a series of text, so as to according to pattern-recognition and meaning
The result of figure evaluation and test establishes the relationship between sentence and entirety.Deep learning modeling of class may be used in user view identification.Needle
To different types of intention, different strategies may be used and respond.
S2, it determines operation system corresponding to the recognition result, and accesses corresponding at least one operation system.
S3, according to the recognition result and the business information extracted from the operation system that is accessed, generate corresponding described chat
The operational order of its content.
Wherein, the operational order generated, for calling at least one operation system.
Further, after the operational order of the corresponding chat content of generation, it is intended that Analysis server can be according to institute
The operational order of generation calls at least one operation system, and performs the business event of at least one operation system.
Specifically, each operation system is docked with being intended to Analysis server, and take notice of and arrange to access and carry in map analysis server
Take the robot of business information.The business event of operation system specifically can be understood as:Operation system is used to handle user's operation
Program event, such as:After user equipment sends order placement information, operation system (can be online shopping platform) places an order
The business event of flow;After information is cancelled an order in user equipment transmission, operation system (can be online shopping platform or visitor
Taking platform) execution of order cancels business event etc. of flow;For another example:It, can after the information for sending predetermined seats in a restaurant in user equipment
According to be intended to Analysis server for user view analysis result, first pass through it is near searching service systematic search user and
Meet the dining room of user demand (chat content that user demand can input user equipment according to user is analyzed to obtain);It recalls
Purchasing by group operation system or intended service system (wherein, can also input according to user and use to the predetermined seats in a restaurant in dining room searched
The chat content of family equipment is analyzed to obtain the information such as specific time for eating meals, institute's menu).
In the present embodiment, it is intended that Analysis server is being answered according to the business information extracted to user equipment transmission
After multiple message, updated chat content is read;And according to the updated chat content, generation is grasped for analog subscriber
The behavioral data of work, and the service node in the operation system is performed according to the behavioral data generated.Such as:To user
Equipment fed back for complete merchandise display pictorial information and for complete product information consulting response message after, after
Continuing can also be according to the chat content newly inputted of user, it is determined whether places an order, plays the operation behaviors such as commodity demonstration video;Also
Can delivery period, place of receipt be changed according to the chat content newly inputted of user;And the chat newly inputted according to user
Content collecting sales promotion information, the product information recommended, preferential Securities and customer accumulate grading information, and are sent to user equipment.
Human-computer interaction implementation method provided in an embodiment of the present invention for customer service, can be by being known by chat content
Other user view determines corresponding operation system, and the enquirement of user is converted into the identifiable operational order of system, and adjust
With each operation system to reach the business event that can be provided in current operation system.Relative in the prior art based on preset
The response mode of " problem base ", the recognition result the present invention is based on user view carry out response, and recall automatically meet user work as
The business of preceding demand, and task, which is reached, to be realized to the subsequent business event of each operation system by chat robots.Not only improve
The responsibilities of chat robots, but also capax negotii is provided, expand application range.
In the present embodiment, it is intended that Analysis server is additionally operable to be sent out to the user equipment according to the business information extracted
Send answer message.Wherein, the mode for sending answer message to the user equipment according to the business information extracted specifically includes:
Extract answer model from the knowledge base, and screened from the business information according to the answer model obtain it is to be feedback
Information, and generate response message according to the information to be feedback.Machine learning and deep learning model may be used, such as:
Classify to problem input by user.And the alternative answer to being extracted is sorted.To subproblem type (as chatted), pass through
The production model generation of deep learning is replied.
Wherein, it specifically includes:
Answer model is extracted from the knowledge base, and is screened and obtained from the business information according to the answer model
Information to be feedback;Response message is generated according to the recognition result and the information to be feedback, is disappeared to the response generated
Breath is screened to obtain answer message, and send answer message to the user equipment
Wherein, described screened from the business information according to the answer model obtains information to be feedback, including:
The answer model of corresponding online transaction platform is extracted from the knowledge base, and is read from the online transaction platform
Business information.Answer model according to the corresponding online transaction platform again inputs the chat content of user equipment for user,
Screening obtains information to be feedback from the business information.
Wherein, read business information includes:For completing merchandise display, shopping guide, product information consulting, purchase decision
Support, performance comparison, order generation, payment flow, logistic track, user feedback and follow-up customer service business information.It is described to treat
Content in feedack with obtaining specific field from the character included by the chat content, has what is mutually met
Part.Wherein, the content in the information to be feedback, with obtaining designated word from the character included by the chat content
Section has the part mutually met.Such as:The chat content that user inputs user equipment has included representing some commodity
Inquiry operation, and the text message for the operation that place an order;Then according to answer model, filter out and be used for from the business information
The pictorial information of merchandise display is completed, for completing the response message of product information consulting, for completing order generation operation, branch
It pays the operation information of flow and display carries out the logistics information of logistic track state.
Further, in the present embodiment, it is intended that Analysis server is also according to the recognition result, from the knowledge base
Reply data is read, and inquiry obtains response message in the reply data read.According to preset ordering rule to inquiry
Obtained response message is ranked up, and obtains the answer message according to ranking results.Wherein, the type of the reply data
It includes at least:Document data, audio data and video data, the response message specified include:Text message, audio-frequency information
Or video information.
Human-computer interaction implementation method provided in an embodiment of the present invention for customer service, by being identified by chat content
User view, determine corresponding operation system, and from operation system extract business information and be used as to the user equipment and send out
Send answer message.Relative to the response mode for being based on preset " problem base " in the prior art, the present invention is based on user views
Recognition result carries out response, and recalls the business for meeting user's current demand automatically.Due to realizing dynamic, obtaining phase in time
The business information of pass carries out human-computer interaction with chat robots, to reach the business event that can be provided in current operation system,
Such as:It is relevant in operation system by obtaining dynamically, in time after commodity retrieval/browsing is carried out by chat robots
Business information can also carry out commercial product recommending, and articles storage places an order, order inquiries tracking, logistics enquiring tracking, and customer service is complained
Business operations are waited, so as in finishing service system, reach (task completion) for the task of each business event, therefore,
Relative to existing chat robots, the present embodiment can also be by chat robots to each operation system subsequent business event reality
Current task is reached, and so as to improve the responsibility of chat robots, expands application range.
In the present embodiment, it can be arranged for different operation systems and respectively arrange corresponding robot, so as to logical
The robot for crossing corresponding operation system extracts business information according to the recognition result from corresponding operation system.For each
Operation system is established the concrete mode of the business information of the operation system of connection by the extraction of intention Analysis server, including:
According to the recognition result, reply data is read from knowledge base, and inquired in the reply data read
To response message.Wherein, the type of the reply data includes at least:Document data, audio data and video data, the finger
Fixed response message includes:Text message, audio-frequency information or video information.Such as:It is intended to the inspection being configured in Analysis server
Rope module includes:The initial query of document retrieval module, mainly Utilizing question analysis generation, retrieves correlation from document library
Document, wherein, the vector space model based on TF*IDF, language model may be used in algorithm;Paragraph or Sentence Search module,
By the quadratic search on the basis of file retrieval result, the document piece for may including problem answers is extracted from document
Section.
The response message obtained later according to preset ordering rule to inquiry is ranked up, and obtained according to ranking results
The answer message.
It can specifically be calculated according to response message and the correlation and similitude identified between obtained specific field
Obtain information distance.And it is ranked up according to the response message that described information distance obtains inquiry.Such as:It can in the present embodiment
To use two stage paragraph sort algorithm, N number of initial paragraph is returned first with the retrieval of IR engines, wherein N be one compared with
Big number, and sorted using relevance ranking algorithm to them.Then selection comes the n (n of foremost<N) a paragraph, and profit
It is resequenced with sort algorithm to this n paragraph.For it is given the problem of qi and its corresponding paragraph catalogue P=
{(pi,j,wi,j), wherein pi,jJ-th of paragraph of problem of representation qi, wi,jRepresent pi,jAbout qiRelevance score, these are related
It is related with the different characteristic of problem to spend score.Then (qi, P) as input feature vector come training machine study sort algorithm.Again
A variety of basic ranking functions are combined using RankBoost, using the ranking results of each basic label function, and in these knots
Learn a best ranking functions on the basis of fruit.
In the present embodiment, using natural language analysis technology parsing sentence, it can also pass through on the basis of retrieval result
The technologies such as pattern match and information extraction extract answer specific field from candidate paragraph or sentence.Such as:Problem " Who
is the president of United States”、“Geoge W.Bush is the president of United
States.”.If template " A_is Q_ ", wherein A_ represents answer part, Q_ problem of representation part, then with the template and above-mentioned
Sentence matches, it is possible to obtain the answer A_ of problem.
In the present embodiment, it is intended that Analysis server can input the chat content of user equipment with user in real, and
The recognition result of user view is obtained according to the chat content.Such as:User is received each time inputs the new of user equipment
Chat messages when, it is intended that Analysis server obtains newest chat messages, and refreshes active user and input chatting for user equipment
Its content, wherein, active user input user equipment chat content included user input user equipment new chat disappear
Breath and the existing chat messages before the new chat messages are received.
If determining the recognition result again according to Analysis server is intended to, analysis obtains need switching currently corresponding
Operation system determines to need to access multiple operation systems, then accesses newest determining operation system.Later, it is intended that analysis
Server is according to the recognition result and the business information extracted from the operation system accessed, the corresponding chat content of generation
Operational order.
And in the present embodiment, it is intended that while the operational order of the corresponding chat content of Analysis server generation,
Can also response message be generated according to the recognition result and the information to be feedback simultaneously, to the response message that is generated into
Row screening obtains answer message, and sends answer message to the user equipment.Wherein, the response message fed back can wrap
It includes:The prompting message of the implementing result of the operational order of the correspondence chat content and corresponding next operating procedure, than
Such as:Lower single operation includes the operating procedure of the confirmation type of merchandise-confirmation commodity amount-payment affirmation:
It is intended to the chat content that Analysis server inputs user equipment according to user, it is determined that the required confirmation to place an order
The type of merchandise is then intended to Analysis server generation for confirming the response message of commodity amount to user:" required purchase please be input
Quantity ";
User according to the prompting of response message, input needed for after the quantity bought, it is intended that Analysis server generation is corresponding true
Recognize the operational order of commodity amount, and send " to ask to the response message of user's confirmation payment behavior to user equipment again
The verification amount of money simultaneously confirms this payment ";
User is according to the prompting of response message, the word or character of input expression " confirmation ", it is intended that Analysis server is given birth to
Into the operational order of corresponding payment affirmation, so as to complete lower single operating process of this commodity.
In the present embodiment, it can also further optimize, update above-mentioned knowledge base, for example further include:To from each industry
It is extracted in business system, is clustered to obtain cluster set, and store to institute according to each service node of the operation system in institute source
It states in knowledge base.Wherein, knowledge base (Knowledge Base) can include:The knowledge data of diversified forms, such as:It is various pure
The knowledge collection of text, various specialized databases, various regular meeting session data, the session flow of various definition, response rule etc..
After corresponding at least one operation system is accessed, the cluster set of corresponding at least one operation system is established
The access interface of conjunction, and pass through the access interface, the business information of each service node in the operation system is read, and make
For the response message.
Wherein it is possible to by training algorithm and historical record, it can be by frequent extraction, application in operation system to deriving from
Carry out cluster, obtain cluster set, and provide the access interface of corresponding cluster set so that be intended to Analysis server and pass through institute
Access interface is stated, the business information of each service node in the operation system is read, and be used as the response message, so as to real
Showed for each operation system can Dynamic Maintenance one cluster set, so as to using cluster set to user initiate
Session carry out rapid answer.Relative to the scheme for carrying out rapid answer in existing scheme by default, fixed " problem base ",
The present embodiment can dynamically adjust the cluster set as " problem base ".
The embodiment of the present invention also provides a kind of human-computer interaction for customer service as shown in Figure 1 and realizes system, wherein
Including:
The intention Analysis server, for obtaining the chat content that user inputs user equipment, and according to the chat
The recognition result of content obtaining user view.And determine the operation system corresponding to the recognition result, and described in access at least
One operation system.Wherein, user equipment for receiving chat content input by user, and chat content is analyzed to intention
Server is sent.
The operation system, for sending business information to the intention Analysis server.
The intention Analysis server is additionally operable to according to the recognition result and the industry extracted from the operation system accessed
Be engaged in information, the operational order of the corresponding chat content of generation, wherein, the operational order generated, for call it is described at least
One operation system.And according to the operational order generated, call at least one operation system.
The operation system is additionally operable to the operational order according to the intention Analysis server generation, performs business event.
In the present embodiment, the intention Analysis server, specifically for from the character included by the chat content
Specific field is obtained, and identifies the information that the specific field includes, the specific field is the character field comprising specify information.
And according to obtained specific field is identified, determine the semanteme and structure of each character field in the chat content.Further according to determining
The chat content in each character field semanteme and structure, the corresponding intention class of the chat content is determined by disaggregated model
Not, and the recognition result is obtained.
In the present embodiment, the intention Analysis server is additionally operable to currently input the chat of user equipment according to user
Content obtains the recognition result that active user is intended to.And the recognition result being intended to according to the active user, update the operation
Instruction.
In the present embodiment, the intention Analysis server, specifically for the identification knot being intended to according to the active user
Fruit switches the operation system of current accessed, and the business information extracted from the operation system after switching.And according to switching after
The business information extracted in operation system updates the operational order.Alternatively, the identification knot being intended to according to the active user
Fruit from the operation system of current accessed, redefines operational order.
Further, the intention Analysis server is additionally operable to extract answer model from the knowledge base, and according to institute
It states answer model and is screened from the business information and obtain information to be feedback.And according to the recognition result and described to be feedback
Information generation response message, screened to obtain answer message to the response message generated, and send out to the user equipment
Send answer message.
Wherein, the intention Analysis server, specifically for extracting corresponding online transaction platform from the knowledge base
Answer model, and business information is read from the online transaction platform, wherein, read business information includes:For completing
Merchandise display, shopping guide, product information consulting, purchase decision support, performance comparison, order generation, payment flow, logistic track,
User feedback and the business information of follow-up customer service.It is defeated for user and according to the answer model of the corresponding online transaction platform
The chat content of access customer equipment screens from the business information and obtains information to be feedback, wherein, the letter to be feedback
Content in breath with obtaining specific field from the character included by the chat content, has the part mutually met.
In the present embodiment, the intention Analysis server is additionally operable to according to the recognition result, from the knowledge base
Reply data is read, and inquiry obtains response message in the reply data read, the type of the reply data is at least wrapped
It includes:Document data, audio data and video data, the response message specified include:Text message, audio-frequency information or video
Information.And the obtained response message of inquiry is ranked up according to preset ordering rule, and obtains according to ranking results described
Answer message.
In the present embodiment, it further includes:Database Systems, for receiving the business extracted from each operation system letter
Breath, and clustered to obtain cluster set, and store to the database according to each service node of the operation system in institute source
In the knowledge base established in system.
The intention Analysis server is additionally operable to after corresponding at least one operation system is accessed, and establishes corresponding institute
The access interface of the cluster set of at least one operation system is stated, and passes through the access interface, is read in the operation system
Each service node business information, and be used as the response message.
In the present embodiment, the intention Analysis server, specifically for identifying what is obtained with described according to response message
Correlation and similitude between specific field, are calculated information distance.And inquiry is obtained according to described information distance
Response message is ranked up.
The embodiment of the present invention also provides a kind of human-computer interaction for customer service as shown in Figure 3 and realizes system, specifically
It may operate on intention Analysis server as shown in Figure 1, including:
Analysis module for obtaining the chat content that user inputs user equipment, and is obtained according to the chat content and is used
The recognition result that family is intended to;It is additionally operable to determine the operation system corresponding to the recognition result, and access corresponding at least one
A operation system;
Module is retrieved, for according to the chat content, extraction to establish the business information of the operation system of connection;
Responder module, for sending answer message to the user equipment according to the business information extracted.
Wherein, the analysis module, specifically for obtaining specific field from the character included by the chat content, and
Identify the information that the specific field includes, the specific field is the character field comprising specify information;And it is obtained according to identification
Specific field, determine the semanteme and structure of each character field in the chat content;Further according to the identified chat content
In each character field semanteme and structure, the corresponding intention classification of the chat content is determined by disaggregated model, and obtain described
Recognition result;
The retrieval module, specifically for according to the recognition result, reply data being read from knowledge base, and reading
To reply data in inquiry obtain response message, the type of the reply data includes at least:Document data, audio data and
Video data, the response message specified include:Text message, audio-frequency information or video information;And according to preset sequence
The response message that rule obtains inquiry is ranked up, and obtain the answer message, the preset row according to ranking results
Sequence rule includes:The response message obtained according to described information distance to inquiry is ranked up, and described information distance is according to response
Information is obtained with the correlation identified between obtained specific field and Similarity measures.
Further, as shown in Figure 4, it can also include:
Knowledge base management module, for being extracted from each operation system, according to each industry of the operation system in institute source
Business node is clustered to obtain cluster set, and store into knowledge base;
The retrieval module, specifically for after corresponding at least one operation system is accessed, establishing described in corresponding to extremely
The access interface of the cluster set of a few operation system, and pass through the access interface, it reads each in the operation system
The business information of service node, and it is used as the response message;
Further, as shown in Figure 5, it can also include:
Model management module, for extracting answer model from the knowledge base, and;
The responder module, specifically for screened from the business information according to the answer model obtain it is to be feedback
Information, and generate response message according to the information to be feedback.
System is realized in human-computer interaction provided in an embodiment of the present invention for customer service, can be by being known by chat content
Other user view determines corresponding operation system, and the enquirement of user is converted into the identifiable operational order of system, and adjust
With each operation system to reach the business event that can be provided in current operation system.Relative in the prior art based on preset
The response mode of " problem base ", the recognition result the present invention is based on user view carry out response, and recall automatically meet user work as
The business of preceding demand, and task, which is reached, to be realized to the subsequent business event of each operation system by chat robots.Not only improve
The responsibilities of chat robots, but also capax negotii is provided, expand application range.Also by being known by chat content
Other user view determines corresponding operation system, and extracts business information from operation system and be used as to the user equipment
Send answer message.Relative to the response mode for being based on preset " problem base " in the prior art, the present invention is based on user views
Recognition result carry out response, and recall the business for meeting user's current demand automatically.Due to realizing dynamic, obtaining in time
Relevant business information carries out human-computer interaction, to reach the business thing that can be provided in current operation system with chat robots
Part, such as:After commodity retrieval/browsing is carried out by chat robots, by dynamic, correlation in operation system is obtained in time
Business information, commercial product recommending can also be carried out, articles storage places an order, order inquiries tracking, logistics enquiring tracking, customer service throw
It tells and waits business operations, so as in finishing service system, reach (task completion) for the task of each business event, because
This, relative to existing chat robots, the present embodiment can also by chat robots to each operation system subsequent business thing
Part realizes that task is reached, and so as to improve the responsibility of chat robots, expands application range.
Each embodiment in this specification is described by the way of progressive, identical similar portion between each embodiment
Point just to refer each other, and the highlights of each of the examples are difference from other examples.Especially for equipment reality
For applying example, since it is substantially similar to embodiment of the method, so describing fairly simple, related part is referring to embodiment of the method
Part explanation.The above description is merely a specific embodiment, but protection scope of the present invention is not limited to
This, any one skilled in the art in the technical scope disclosed by the present invention, the variation that can readily occur in or replaces
It changes, should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with the protection model of claim
Subject to enclosing.
Claims (17)
1. a kind of human-computer interaction implementation method for customer service, which is characterized in that including:
The chat content that user inputs user equipment is obtained, and the recognition result of user view is obtained according to the chat content;
It determines the operation system corresponding to the recognition result, and accesses at least one operation system;
According to the recognition result and the business information extracted from the operation system accessed, the corresponding chat content of generation
Operational order, wherein, the operational order generated, for calling at least one operation system.
2. it according to the method described in claim 1, it is characterized in that, further includes:
According to the operational order generated, at least one operation system is called, and perform at least one operation system
Business event.
3. the according to the method described in claim 1, it is characterized in that, knowledge that user view is obtained according to the chat content
Not as a result, including:
Specific field is obtained from the character included by the chat content, and identifies the information that the specific field includes, institute
Specific field is stated as the character field comprising specify information;
According to the specific field that identification obtains, the semanteme and structure of each character field in the chat content are determined;
According to the semanteme and structure of each character field in the identified chat content, determined in the chat by disaggregated model
Hold corresponding intention classification, and obtain the recognition result.
4. it according to the method described in claim 1, it is characterized in that, further includes:
The chat content of user equipment is currently inputted according to user, obtains the recognition result that active user is intended to;
According to the recognition result that the active user is intended to, the operational order is updated.
5. according to the method described in claim 2, it is characterized in that, it is described according to the active user be intended to recognition result,
The operational order is updated, including:
According to the recognition result that the active user is intended to, switch the operation system of current accessed, and from the business system after switching
The business information extracted in system;And according to the business information extracted in the operation system after switching, update the operational order;
Alternatively, according to the recognition result that the active user is intended to, from the operation system of current accessed, redefine operation and refer to
It enables.
6. it according to the method described in claim 1, it is characterized in that, further includes:
Extract answer model from knowledge base, and screened from the business information according to the answer model obtain it is to be feedback
Information;
Response message is generated according to the recognition result and the information to be feedback, the response message generated is screened
Answer message is obtained, and answer message is sent to the user equipment.
7. according to the method described in claim 6, it is characterized in that, it is described according to the answer model from the business information
Screening obtains information to be feedback, including:
The answer model of corresponding online transaction platform is extracted from the knowledge base, and business is read from the online transaction platform
Information, wherein, read business information includes:For completing merchandise display, shopping guide, product information consulting, purchase decision branch
Hold, the generation of performance comparison, order, payment flow, logistic track, user feedback and follow-up customer service business information;
According to the answer model of the corresponding online transaction platform, the chat content of user equipment is inputted for user, from described
Screening obtains information to be feedback in business information, wherein, the content in the information to be feedback, and from the chat content
Specific field is obtained in included character, there is the part mutually met.
8. it according to the method described in claim 6, it is characterized in that, further includes:
According to the recognition result, reply data is read from the knowledge base, and inquired in the reply data read
To response message, the type of the reply data includes at least:Document data, audio data and video data, it is described to specify
Response message includes:Text message, audio-frequency information or video information;
The response message obtained according to preset ordering rule to inquiry is ranked up, and obtain the answer according to ranking results
Message.
9. it according to the method described in claim 8, it is characterized in that, further includes:
To the business information extracted from each operation system, clustered according to each service node of the operation system in institute source
Cluster set is obtained, and is stored into the knowledge base;
After corresponding at least one operation system is accessed, the cluster set of corresponding at least one operation system is established
Access interface, and pass through the access interface, the business information of each service node in the operation system is read, and as institute
State response message.
10. according to the method described in claim 8, it is characterized in that, described obtain inquiry according to preset ordering rule
Response message is ranked up, including:
According to response message and the correlation and similitude identified between obtained specific field, be calculated information away from
From;
The response message obtained according to described information distance to inquiry is ranked up.
11. system is realized in a kind of human-computer interaction for customer service, which is characterized in that including:
The intention Analysis server, for obtaining the chat content that user inputs user equipment, and according to the chat content
Obtain the recognition result of user view;And determine the operation system corresponding to the recognition result, and access described at least one
Operation system;
The operation system, for sending business information to the intention Analysis server;
The intention Analysis server is additionally operable to be believed according to the recognition result and the business extracted from the operation system accessed
Breath, the operational order of the corresponding chat content of generation, wherein, the operational order generated is described at least one for calling
Operation system;And according to the operational order generated, call at least one operation system;
The operation system is additionally operable to the operational order according to the intention Analysis server generation, performs business event.
12. system according to claim 11, which is characterized in that the intention Analysis server, specifically for from described
Specific field is obtained in character included by chat content, and identifies the information that the specific field includes, the specific field
For the character field comprising specify information;And the specific field obtained according to identification, determine each character field in the chat content
Semantic and structure;Further according to the semanteme and structure of each character field in the identified chat content, determined by disaggregated model
The corresponding intention classification of the chat content, and obtain the recognition result.
13. system according to claim 11, which is characterized in that the intention Analysis server is additionally operable to according to user
The chat content of current input user equipment, obtains the recognition result that active user is intended to;And it is intended to according to the active user
Recognition result, update the operational order.
14. system according to claim 12, which is characterized in that the intention Analysis server, specifically for according to institute
The recognition result of active user's intention is stated, switches the operation system of current accessed, and extracted from the operation system after switching
Business information;And according to the business information extracted in the operation system after switching, update the operational order;Alternatively, according to institute
The recognition result of active user's intention is stated, from the operation system of current accessed, redefines operational order.
15. system according to claim 11, which is characterized in that the intention Analysis server is additionally operable to from knowledge base
Middle extraction answer model, and screened from the business information according to the answer model and obtain information to be feedback;And according to
The recognition result and the information generation response message to be feedback, screen the response message generated and are replied
Message, and send answer message to the user equipment.
16. system according to claim 15, which is characterized in that the intention Analysis server, specifically for from described
The answer model of corresponding online transaction platform is extracted in knowledge base, and business information is read from the online transaction platform, wherein,
Read business information includes:For completing merchandise display, shopping guide, product information consulting, purchase decision support, performance pair
Than the business information of, order generation, payment flow, logistic track, user feedback and follow-up customer service;It is and online according to the correspondence
The answer model of transaction platform inputs the chat content of user equipment for user, screens and treated from the business information
Feedack, wherein, the content in the information to be feedback, with being referred to from the character included by the chat content
Determine field, there is the part mutually met.
17. system according to claim 15, which is characterized in that the intention Analysis server is additionally operable to according to
Recognition result reads reply data, and inquiry obtains response message in the reply data read from the knowledge base, institute
The type for stating reply data includes at least:Document data, audio data and video data, the response message specified include:
Text message, audio-frequency information or video information;And be ranked up according to the response message that preset ordering rule obtains inquiry,
And obtain the answer message according to ranking results.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611197153.5A CN108228559A (en) | 2016-12-22 | 2016-12-22 | A kind of human-computer interaction realization method and system for customer service |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611197153.5A CN108228559A (en) | 2016-12-22 | 2016-12-22 | A kind of human-computer interaction realization method and system for customer service |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108228559A true CN108228559A (en) | 2018-06-29 |
Family
ID=62656830
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611197153.5A Pending CN108228559A (en) | 2016-12-22 | 2016-12-22 | A kind of human-computer interaction realization method and system for customer service |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108228559A (en) |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109346078A (en) * | 2018-11-09 | 2019-02-15 | 泰康保险集团股份有限公司 | Voice interactive method, device and electronic equipment, computer-readable medium |
CN109582760A (en) * | 2018-08-06 | 2019-04-05 | 北京马上慧科技术有限公司 | A kind of response flow engine system applied in Intelligent dialogue system |
CN109710941A (en) * | 2018-12-29 | 2019-05-03 | 上海点融信息科技有限责任公司 | User's intension recognizing method and device based on artificial intelligence |
CN110516035A (en) * | 2019-07-05 | 2019-11-29 | 同济大学 | A kind of man-machine interaction method and system of mixing module |
CN110769312A (en) * | 2019-11-07 | 2020-02-07 | 三角兽(北京)科技有限公司 | Method and device for recommending information in live broadcast application |
CN111091826A (en) * | 2019-12-13 | 2020-05-01 | 中博信息技术研究院有限公司 | Intelligent voice robot system based on deep learning and finite-state machine |
WO2020119031A1 (en) * | 2018-12-11 | 2020-06-18 | 平安科技(深圳)有限公司 | Deep learning-based question and answer feedback method, device, apparatus, and storage medium |
CN111782792A (en) * | 2020-08-05 | 2020-10-16 | 支付宝(杭州)信息技术有限公司 | Method and apparatus for information processing |
CN112104541A (en) * | 2019-06-17 | 2020-12-18 | 夏普株式会社 | Network system, server, information processing method, and recording medium |
CN112699228A (en) * | 2020-12-31 | 2021-04-23 | 青岛海尔科技有限公司 | Service access method, device, electronic equipment and storage medium |
CN113422810A (en) * | 2021-06-02 | 2021-09-21 | 广州三星通信技术研究有限公司 | Method and device for sending information to service provider |
CN113689253A (en) * | 2021-07-09 | 2021-11-23 | 广州华多网络科技有限公司 | Live scene order generation method and corresponding device, equipment and medium thereof |
WO2022012107A1 (en) * | 2020-07-17 | 2022-01-20 | 北京沃东天骏信息技术有限公司 | Method and apparatus for displaying plug-in |
CN115514571A (en) * | 2022-09-26 | 2022-12-23 | 中国电信股份有限公司 | Customer service method and device based on zero trust, electronic equipment and readable storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102376014A (en) * | 2010-08-10 | 2012-03-14 | 中国科学院计算技术研究所 | Response management method supporting diversity of clients and businesses and system thereof |
US20140344788A1 (en) * | 2013-05-15 | 2014-11-20 | Oracle International Corporation | Logic validation and deployment |
CN104951458A (en) * | 2014-03-26 | 2015-09-30 | 华为技术有限公司 | Method and equipment for helping processing based on semantic recognition |
CN105389307A (en) * | 2015-12-02 | 2016-03-09 | 上海智臻智能网络科技股份有限公司 | Statement intention category identification method and apparatus |
CN105721725A (en) * | 2016-02-03 | 2016-06-29 | 北京光年无限科技有限公司 | Customer service oriented question and answer interaction method and system |
-
2016
- 2016-12-22 CN CN201611197153.5A patent/CN108228559A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102376014A (en) * | 2010-08-10 | 2012-03-14 | 中国科学院计算技术研究所 | Response management method supporting diversity of clients and businesses and system thereof |
US20140344788A1 (en) * | 2013-05-15 | 2014-11-20 | Oracle International Corporation | Logic validation and deployment |
CN104951458A (en) * | 2014-03-26 | 2015-09-30 | 华为技术有限公司 | Method and equipment for helping processing based on semantic recognition |
CN105389307A (en) * | 2015-12-02 | 2016-03-09 | 上海智臻智能网络科技股份有限公司 | Statement intention category identification method and apparatus |
CN105721725A (en) * | 2016-02-03 | 2016-06-29 | 北京光年无限科技有限公司 | Customer service oriented question and answer interaction method and system |
Non-Patent Citations (1)
Title |
---|
黎新: "面向问答系统的段落检索技术研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109582760A (en) * | 2018-08-06 | 2019-04-05 | 北京马上慧科技术有限公司 | A kind of response flow engine system applied in Intelligent dialogue system |
CN109346078A (en) * | 2018-11-09 | 2019-02-15 | 泰康保险集团股份有限公司 | Voice interactive method, device and electronic equipment, computer-readable medium |
CN109346078B (en) * | 2018-11-09 | 2021-06-18 | 泰康保险集团股份有限公司 | Voice interaction method and device, electronic equipment and computer readable medium |
WO2020119031A1 (en) * | 2018-12-11 | 2020-06-18 | 平安科技(深圳)有限公司 | Deep learning-based question and answer feedback method, device, apparatus, and storage medium |
CN109710941A (en) * | 2018-12-29 | 2019-05-03 | 上海点融信息科技有限责任公司 | User's intension recognizing method and device based on artificial intelligence |
CN112104541A (en) * | 2019-06-17 | 2020-12-18 | 夏普株式会社 | Network system, server, information processing method, and recording medium |
CN110516035A (en) * | 2019-07-05 | 2019-11-29 | 同济大学 | A kind of man-machine interaction method and system of mixing module |
CN110769312A (en) * | 2019-11-07 | 2020-02-07 | 三角兽(北京)科技有限公司 | Method and device for recommending information in live broadcast application |
CN111091826A (en) * | 2019-12-13 | 2020-05-01 | 中博信息技术研究院有限公司 | Intelligent voice robot system based on deep learning and finite-state machine |
CN111091826B (en) * | 2019-12-13 | 2023-09-01 | 中博信息技术研究院有限公司 | Intelligent voice robot system based on deep learning and finite state machine |
WO2022012107A1 (en) * | 2020-07-17 | 2022-01-20 | 北京沃东天骏信息技术有限公司 | Method and apparatus for displaying plug-in |
CN111782792A (en) * | 2020-08-05 | 2020-10-16 | 支付宝(杭州)信息技术有限公司 | Method and apparatus for information processing |
CN112699228A (en) * | 2020-12-31 | 2021-04-23 | 青岛海尔科技有限公司 | Service access method, device, electronic equipment and storage medium |
CN112699228B (en) * | 2020-12-31 | 2023-07-14 | 青岛海尔科技有限公司 | Service access method, device, electronic equipment and storage medium |
CN113422810A (en) * | 2021-06-02 | 2021-09-21 | 广州三星通信技术研究有限公司 | Method and device for sending information to service provider |
CN113689253A (en) * | 2021-07-09 | 2021-11-23 | 广州华多网络科技有限公司 | Live scene order generation method and corresponding device, equipment and medium thereof |
CN115514571A (en) * | 2022-09-26 | 2022-12-23 | 中国电信股份有限公司 | Customer service method and device based on zero trust, electronic equipment and readable storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108228559A (en) | A kind of human-computer interaction realization method and system for customer service | |
KR101938790B1 (en) | Auto response server device, terminal device, reply system, reply method, and program | |
US20220006761A1 (en) | Systems and processes for operating and training a text-based chatbot | |
US11334635B2 (en) | Domain specific natural language understanding of customer intent in self-help | |
US10755317B2 (en) | Managing a set of offers using a dialogue | |
CN108536852A (en) | Question and answer exchange method and device, computer equipment and computer readable storage medium | |
CN109690602A (en) | Products Show is provided in automatic chatting | |
CN107636648A (en) | Response is constructed based on mood mark | |
CN104969173A (en) | Method for adaptive conversation state management with filtering operators applied dynamically as part of a conversational interface | |
JP2003330948A (en) | Device and method for evaluating web page | |
US11632341B2 (en) | Enabling communication with uniquely identifiable objects | |
CN109582798A (en) | Automatic question-answering method, system and equipment | |
CN109388697A (en) | Handle the method, apparatus and electric terminal of question and answer corpus | |
CN107291900A (en) | Feedback of the information and tracking system | |
Vivek et al. | A novel technique for user decision prediction and assistance using machine learning and NLP: A model to transform the E-commerce system | |
Kumar et al. | Chatbot in Python | |
Daniati et al. | Decision Making Framework Based On Sentiment Analysis in Twitter Using SAW and Machine Learning Approach | |
US11966698B2 (en) | System and method for automatically tagging customer messages using artificial intelligence models | |
Iwai et al. | A help desk support system with filtering and reusing e-mails | |
US20210117972A1 (en) | Rule-based messaging and electronic communication system | |
US20230024052A1 (en) | Techniques for managing softphone repositories and establishing communication channels | |
CN110428342A (en) | Data recovery method, server, customer side and storage medium | |
CN109829744A (en) | Consultation method, device, electronic equipment and medium based on natural language processing | |
CN108491506A (en) | Method for pushing problem answers combination | |
Prasomphan | Using chatbot in trading system for small and medium enterprise (smes) by Convolution Neural Network Technique |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | 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: 20180629 |