CN109919671A - The high efficiency smart client service method and system of a kind of Electronic Commerce field customer orientation - Google Patents
The high efficiency smart client service method and system of a kind of Electronic Commerce field customer orientation Download PDFInfo
- Publication number
- CN109919671A CN109919671A CN201910146974.3A CN201910146974A CN109919671A CN 109919671 A CN109919671 A CN 109919671A CN 201910146974 A CN201910146974 A CN 201910146974A CN 109919671 A CN109919671 A CN 109919671A
- Authority
- CN
- China
- Prior art keywords
- service
- client
- module
- quality
- efficiency
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 72
- 230000004044 response Effects 0.000 claims abstract description 53
- 230000036651 mood Effects 0.000 claims abstract description 40
- 230000008569 process Effects 0.000 claims abstract description 32
- 238000011156 evaluation Methods 0.000 claims abstract description 23
- 238000013139 quantization Methods 0.000 claims abstract description 9
- 238000001514 detection method Methods 0.000 claims description 38
- 238000012544 monitoring process Methods 0.000 claims description 19
- 238000012360 testing method Methods 0.000 claims description 14
- 230000002996 emotional effect Effects 0.000 claims description 12
- 238000013528 artificial neural network Methods 0.000 claims description 8
- 230000007474 system interaction Effects 0.000 claims description 3
- 238000007726 management method Methods 0.000 description 24
- 230000000694 effects Effects 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 238000009223 counseling Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000013507 mapping Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000005192 partition Methods 0.000 description 3
- 230000007704 transition Effects 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000008451 emotion Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 239000013589 supplement Substances 0.000 description 2
- 238000003325 tomography Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to a kind of high efficiency smart client service method of Electronic Commerce field customer orientation and systems.The present invention is based on the efficiency and quality problems for solving the intelligent client service system under electric business scene, reformulate the service quality of consulting response and the index of efficiency of service, integrated service quality and service efficiency evaluation standard customer-centric are defined, propose the new model and process of robot service and man-machine collaboration: first stage is simple machine service;Second stage is machine service, artificial to supervise;Three phases are manual services, while machine assists.The present invention helps to provide preferably consulting answer service experience for client, man-machine handover seamless and it is smooth;The mood of client is paid attention to and is protected, and can cut manual service in time, obtains more intimate quality service;Can real-time quantization service level, including service quality and efficiency, and timely feedback and make corresponding adjustment to machine and artificial customer service.
Description
Technical field
The invention belongs to artificial intelligence, big data and field of information processing, focus on e-commerce industry, and in particular to one
The high efficiency smart client service method and system of kind Electronic Commerce field customer orientation.
Background technique
The needs of e-commerce to flourish along with consumption upgrading, consumer participate in a variety of manners more and more
On to line in consumption, while the requirement to the service experience during consumption is also increasingly to improve.Correspondingly, to winning more
The time of more consumers and loyalty, businessman will mention from multiple links such as commodity, logistics, dispatching and customer service to consumer
For optimal shopping experience.Wherein, client's consulting and service is through the core of entire process of consumption, including pre-sales, mid-sales and
After sale.Consumer can seek advice from any problem, the opinion etc. in process of consumption by customer service channel substantially, complain and
It solves.Therefore, substantially any household appliances manufacturer platform can provide the function of customer service consulting and service, quotient thereon all for consumer
Family and subject of operation also can arrange special contact staff to receive client, answer and solution customer issue.
Artificial customer service is still main customer service ability form at present, and the chat tool of instant messaging is often most important
Means.But more and more electric business platforms and brand side start input research and apply customer service robot, it is therefore an objective to utilize visitor
Robot is taken to cope with a large amount of simple consultation and advices, to save the cost of artificial customer service investment.The customer service of current several mainstreams
Robotic scenarios have:
1. the customer service robot of guiding: the mode of similar multistage navigating directory guides user's positioning to ask step by step
The problem of inscribing, and finally presetting answer is presented to user, or is transferred to artificial customer service.This method implements simple straight
It connects, the commodity high for a small amount of standardization level are relatively applicable in.Maximum problem is exactly that user cannot actively ask a question, it is necessary to passive
Ground receives guidance, and efficiency and experience are all poor.
2. the customer service robot of class search engine formula: user can actively send out problem, and customer service robot is known in rear end
Know in library and carry out full-text search, maximally related content retrieval can will be putd question to come out with user, and be presented to the user.This kind of scheme
Challenge is how precisely to identify the intention of client, else if if the result retrieved is excessive, then can be difficult to
Family is presented most accurately as a result, user experience is caused to be deteriorated.
3. the heuristic customer service robot of knowledge based map: i.e. the relevant data organization in field at network structure
Knowledge mapping, including scene be applicable in language message (such as shopping idiomatic expression method, idiom), brand and commodity number
According to, the fast delivery data of logistics, the relevant common sense of platform and idiom etc..Such knowledge mapping can support customer service robot fast
Speed accurately identifies and positions user's needs, and " closed domain " of so-called dialogue is entered with most succinct step guidance user, is realized
Efficient service.Session log of the advantage of this scheme without a large number of users and customer service, it is only necessary to be based on existing neck
Numeric field data can be realized " cold start-up " and construct customer service robot.But disadvantage is often that the scope of knowledge mapping is difficult to have covered
Whole customer service demand needs continuous fill rule and data, also to prevent rule conflict, the cost of upgrading and maintenance compared with
It is high.
4. the customer service robot based on deep learning: the core of this kind of scheme is all to a large amount of user and customer service dialogue note
The main points of view under different scenes may be implemented using different deep neural network models in the training that record carries out deep learning
Or the ability of consulting intention assessment.The problem of this scheme is, first, since it is desired that a large amount of corpus is for training, so
The cost of the building of system initial stage and starting can be relatively high;Second, single model is difficult suitable for multiple scenes or a scene
Multiple links, therefore applicable surface is relatively narrow.
All there is a common modes for above scheme, that is, is first replaced manually, then existing completely by robot
Artificial customer service is handed to again when question and answer can not be handled in the existing System of Logic of robot to be handled, and next just no longer
Pay close attention to the case where question and answer are handled on artificial customer service line.This is a kind of typical two stages service mode.Between two stages often
The phenomenon that there is service " partitions ", artificial customer service also needs to look back when taking over consulting and analysis robot is in the first rank
The case where section response, or even problem is looked back with user again.User experience has thus been seriously affected, customer service effect is reduced
Rate.It is crucial that the technology of recognition capability of the powerful query capability and deep learning model of knowledge mapping in specific link is excellent
Gesture also can not be implemented and be embodied.
For example, patent document CN107590159A, publication date 2018.01.16, it is artificial to disclose a kind of robot customer service turn
The method of customer service, comprising: obtain validity feature from robot customer service and at least one wheel session of user;Validity feature is inputted
Confidence assessment models obtain the current confidence assessed value of robot customer service and user conversation;The confidence assessment models are using mark
Note has validity feature and the session sample of the suitable robot customer service and user manually put out to be trained, and described is suitable for artificial out
Point is the appropriate time point with artificial customer service instead of robot customer service;When current confidence assessed value meets and makes artificial condition in advance,
User is transferred artificial customer service.The method of the patent document is still typical two stages service mode.
It has no Electronic Commerce field at present, defines integrated service quality and efficiency of service customer-centric
Evaluation criteria, three stage man-machine collaboration customer service modes, realizes high customer satisfaction, efficient client service method and system.
Summary of the invention
The purpose of the present invention is aiming at the shortcomings in the prior art, provide a kind of Electronic Commerce field customer orientation
High efficiency smart client service method.
Another purpose of the invention is to provide a kind of high efficiency smart customer service system of Electronic Commerce field customer orientation
System.
To realize above-mentioned first purpose, the technical solution adopted by the present invention is that:
A kind of high efficiency smart client service method of Electronic Commerce field customer orientation, comprising the following steps:
Step S1, in the first stage, the chat that context of dialogue management module is responsible for receiving client's dialog interface module disappear
Breath calls intention assessment module and mood detection module to analyze the enquirement of client, and then search knowledge base obtains crucial
Information simultaneously utilizes response template generation answer statement, and answer statement is replied to client finally by client's dialog interface module;
Step S2, if context of dialogue management module detects customer anger exception or intention assessment unstable result, that
Second stage is entered, artificial customer service background module is started, sends contact staff one end for response message, into customer service people
In the supervision stage of member, when the supervision stage for entering contact staff, the contextual information of dialogue can be presented on the auxiliary at customer service interface
Information area, to help contact staff to quickly understand the process entirely talked with;
Step S3, contact staff choose whether to enter phase III, i.e., artificial customer service according to the emotional state of existing customer
Leading, in the stage of machine auxiliary, in this stage, context of dialogue management module calls knowledge base that contact staff is helped to inquire institute
The knowledge needed is as auxiliary;
Quality and efficiency monitoring module acquire the key message of each round of links in dialog procedure in real time, to clothes
Quality of being engaged in and efficiency of service carry out real-time quantization calculating and assessment;The evaluation criteria of service quality includes three factors: Ke Hushi
The no answer being satisfied with, client receive service during the entire process of whether be it is pleasant, whether response content meets product
Board or the prepared service regulation of businessman;The assessment of efficiency of service is for entire service process, and evaluation criteria includes two
Factor: the round and response speed of dialogue.
As a preference, the key message of the quality and the acquisition of efficiency monitoring module includes: to talk with from client
The timestamp of the sentence of interface module, intention testing result and mood testing result from context of dialogue management module are come
From the timestamp of the consulting response of artificial customer service background module.
It is highly preferred that reply efficiency is calculated by timestamp, detect to obtain client by being intended to detection and mood
Satisfaction.
As another preference, for the intention assessment module using the text of the consulting sentence of client as input, output should
Client expressed by sentence is intended to, it is intended that knowing method for distinguishing is the detection based on deep neural network;The mood detection module
Using the text of the consulting sentence of client as input, the emotional state that the sentence is reflected is exported, the method for mood detection is
Detection based on deep neural network.
As another preference, the quality and the real-time evaluation services quality of efficiency monitoring module and efficiency of service, and and
When feed back to machine and artificial customer service to make corresponding adjustment.
To realize above-mentioned second purpose, the technical solution adopted by the present invention is that:
A kind of high efficiency smart customer service system of Electronic Commerce field customer orientation, comprising:
Client terminal;
Client's dialog interface module: i.e. the interface of terminal client and customer service system interaction, for realizing instant messaging;
Context of dialogue management module: it is responsible for the core business logic and contextual information of management of dialogs system, control three
The propulsion and circulation of phase flow, the core business logic is:
1) context of dialogue management module is responsible for receiving the chat messages of client's dialog interface module in the first stage, calls
Intention assessment module and mood detection module analyze the enquirement of client, and then search knowledge base obtains key message and benefit
With response template generation answer statement, answer statement is replied into client finally by client's dialog interface module;
2) customer anger exception or intention assessment unstable result are detected such as context of dialogue management module, then just into
Enter second stage, start artificial customer service background module, send contact staff one end for response message, into the prison of contact staff
See the stage, when the supervision stage for entering contact staff, the contextual information of dialogue can be presented on the slug area at customer service interface
Domain, to help contact staff to quickly understand the process entirely talked with;
3) contact staff chooses whether to enter the phase III according to the emotional state of existing customer, i.e., artificial customer service is leading,
The stage of machine auxiliary, in this stage, context of dialogue management module call knowledge base to help required for contact staff's inquiry
Knowledge as auxiliary;
Artificial customer service background module: supervision and response consulting are realized for contact staff;
Knowledge base: tissue is used for the core data of response, carries out real-time query;
Mood detection module: whether normal for detecting mood during client questions;
Intention assessment module: for detecting the intention of client questions;
Quality and efficiency monitoring module: believed by acquiring the crucial of each round of links in dialog procedure in real time
Breath carries out real-time quantization calculating and assessment to service quality and efficiency of service;The evaluation criteria of service quality includes three factors:
The answer whether client is satisfied with, client receive service during the entire process of whether be it is pleasant, whether response content
Meet brand or the prepared service regulation of businessman;The assessment of efficiency of service is the evaluation criteria packet for entire service process
Include two factors: the round and response speed of dialogue;
Contact staff's terminal.
As a preference, the key message of the quality and the acquisition of efficiency monitoring module includes: to talk with from client
The timestamp of the sentence of interface module, intention testing result and mood testing result from context of dialogue management module are come
From the timestamp of the consulting response of artificial customer service background module.
It is highly preferred that reply efficiency is calculated by timestamp, detect to obtain client by being intended to detection and mood
Satisfaction.
As another preference, for the intention assessment module using the text of the consulting sentence of client as input, output should
Client expressed by sentence is intended to, it is intended that knowing method for distinguishing is the detection based on deep neural network;The mood detection module
Using the text of the consulting sentence of client as input, the emotional state that the sentence is reflected is exported, the method for mood detection is
Detection based on deep neural network.
As another preference, the quality and the real-time evaluation services quality of efficiency monitoring module and efficiency of service, and and
When feed back to machine and artificial customer service to make corresponding adjustment.
The invention has the advantages that:
1, the present invention is again fixed based on the efficiency and quality problems for solving the intelligent client service system under electric business scene
The service quality of justice consulting response and the index of efficiency of service, define integrated service quality kimonos customer-centric
Business efficiency evaluation standard.The evaluation criteria of service quality includes three factors: the answer whether client is satisfied with, Ke Hu
Receive service during the entire process of whether be it is pleasant, whether response content meets brand or the prepared service regulation of businessman.
The assessment of efficiency of service is including the machine and artificial for entire service process, and evaluation criteria includes two factors: dialogue
Round and response speed.By the emotional change of client in quantization consulting answering, reduce because customer anger problem is coped with
It is complained caused by improper or difference is commented;By using customer service specification specified by brand or businessman as important knowledge rule,
With further hoisting machine people response quality.Facilitate the service level of General Promotion intelligent customer service system.
2, the invention proposes the new models and process of robot service and man-machine collaboration.The mode of man-machine collaboration is by typical case
Two stages develop into three phases: first stage is exactly simple machine service;Second stage is machine service, manually
It supervises;Three phases are manual services, while machine assists.The design of core process in this way can greatly improve biography
The service quality for two stages process of uniting, while guaranteeing very high efficiency of service.It, can in introducing of the second stage as transition stage
To allow manually carry out auxiliary machinery and judged, while buffer time is provided to be transformed into manual service, the experience of such client is just
Can be smoother, two stage " partition " or " tomography " sense is eliminated, realizes robot service and manual service seamless interfacing.
3, the present invention to client be intended to identification and mood detection be using client consulting sentence text as input, this
It is different from conventional Emotion identification mode, but system and method for the invention can be made especially suitable for e-commerce industry.
Generally speaking, the present invention once puts into application, and following technical effect may be implemented: providing better consulting for client
Answer service experience, man-machine handover seamless and it is smooth;The mood of client is paid attention to and is protected, and can cut artificial clothes in time
Business, obtains more intimate quality service;Can real-time quantization service level, including service quality and efficiency of service, and anti-in time
It feeds machine and corresponding adjustment is made in artificial customer service.
Detailed description of the invention
Attached drawing 1 is the flow diagram of high efficiency smart client service method of the present invention.
Attached drawing 2 is the structural block diagram of high efficiency smart customer service system of the present invention.
Specific embodiment
It elaborates with reference to the accompanying drawing to specific embodiment provided by the invention.
Appended drawing reference involved in attached drawing and component part are as follows:
1. 2. client's dialog interface module of client terminal
3. the artificial customer service background module of context of dialogue management module 4.
5. 6. mood detection module of knowledge base
7. 8. mass of intention assessment module and efficiency monitoring module
9. contact staff's terminal
The high efficiency smart client service method of the present invention of embodiment 1
Referring to Figure 1, Fig. 1 is the flow diagram of high efficiency smart client service method of the present invention.The high efficiency smart customer service side
Method the following steps are included:
Step S1, in the first stage, the chat that context of dialogue management module is responsible for receiving client's dialog interface module disappear
Breath calls intention assessment module and mood detection module to analyze the enquirement of client, and then search knowledge base obtains crucial
Information simultaneously utilizes response template generation answer statement, and answer statement is replied to client finally by client's dialog interface module.
Step S2, it is entire to seek advice from response if client can terminate to seek advice from simple two into three-wheel dialogue
Journey terminates;Otherwise context of dialogue management module detects customer anger exception or intention assessment unstable result, then just into
Enter second stage, at this time just start artificial customer service background module, contact staff one end is sent by response message, into customer service
The supervision stage of personnel.
Step S3, if can complete to talk in second stage, entire answering of seeking advice from terminates;Otherwise customer service people
Member can select to enter the phase III according to field conditions, i.e., artificial customer service is leading, the stage of machine auxiliary.In this stage, right
Knowledge required for words context management module can call knowledge base that contact staff is helped to inquire is detailed as auxiliary, such as commodity
Feelings, order information, logistics distribution information, counseling services specification masks of brand, etc..Meanwhile context of dialogue management module
Also mood detection module and intention assessment module can be called to analyze the conversation content on both sides.
Further, when the supervision stage for entering contact staff, the contextual information of dialogue can be presented on customer service interface
Auxiliary information region, to help contact staff to quickly understand the process entirely talked with.Meanwhile contact staff can also be according to current
The emotional state of client comes to decide whether to service sovereignty adapter tube.
Further, the high efficiency smart client service method further includes the steps that quality and efficiency monitoring: by acquiring in real time
The key message of each round of links in dialog procedure, efficiency and quality to dialogue carry out real-time quantization calculating and comment
Estimate.The data of acquisition include the timestamp of the sentence from client's dialog interface module, from context of dialogue management module
It is intended to testing result and mood testing result, the timestamp of the consulting response from artificial customer service background module.Pass through timestamp
Calculating, available reply efficiency detects available customer satisfaction by being intended to detection and mood, so that synthesis obtains
The service level assessment of response consulting.Specifically, the step comprehensive assessment of quality and efficiency monitoring goes out service quality and service effect
Rate.The evaluation criteria of the service quality includes three factors: first is exactly the answer whether client is satisfied with, and second just
Client during the entire process of receiving service whether be it is pleasant, third is exactly whether response content meets brand or businessman institute
The service regulation of formulation.The assessment of the efficiency of service is including machine and artificial, evaluation criteria for entire service process
Including two factors: the round and response speed of dialogue.Preferably, service quality and efficiency of service Real-time Feedback are to machine and people
Work customer service is to make corresponding adjustment.
Further, the knowledge base is used for the core data of response by RDF mode tissue, is carried out using SparQL
Real-time query, the core data mainly include the counseling services rule of commodity details, order information, logistics distribution information, brand
Model template, etc..Knowledge base can constantly be expanded.
It should be noted that high efficiency smart client service method of the invention defines integrated service customer-centric
Quality and service efficiency evaluation standard, so-called integration assessment are the unified examination under man-machine collaboration.From the viewpoint of client,
Whether machine or artificial, requires to be capable of providing good consulting answer service.So the evaluation criteria of service quality is answered
This includes three important factors: first is exactly the answer whether client is satisfied with, and second is exactly that client is receiving service
During the entire process of whether be it is pleasant, third is exactly whether response content meets brand or the prepared service regulation of businessman.
Meanwhile the evaluation criteria of efficiency of service is then including machine and artificial, the round mainly talked with for entire service process
Determine efficiency of service with response speed: round is fewer, and it is higher to respond more fast then efficiency of service.
The invention proposes the new models and process of robot service and man-machine collaboration.The mode of man-machine collaboration is by typical
Two stages develop into three phases: first stage is exactly simple machine service;Second stage is machine service, artificial to supervise
It sees;Three phases are manual services, while machine assists.With traditional two stages mode the difference is that second and third rank
Section.
Second stage is that machine turns an artificial transition stage, and the artificial rank being familiar with and confirm the context of dialogue
Section.The condition that the stage is triggered mainly includes that machine can not identify to accurate stable the intention that user puts question to, can not be quick
User is brought into " closed domain ", user emotion occurs abnormal, etc..When detecting these problems, contact staff will
The case where being connected to chat request and open dialog interface, and real-time monitoring machine and user session.This when, dialogue up and down
Literary information can be presented on the auxiliary information region at customer service interface, to help contact staff to quickly understand the process entirely talked with.Together
When, contact staff can also decide whether that will service sovereignty adapter tube comes according to the emotional state of existing customer.
Phase III is not to be serviced completely by manually, but machine indirect labor realizes service.This is also meaned that
Artificial adapter tube comes after service, and machine can serve as the effect of one " brain truster ", because some task equipments can more be increased to do
Effect, to help contact staff's high-speed decision, quick response.For example, the positioning of inquiry and key concept to commodity information database, right
The inquiry of brand marketers customer service specification and the assessment of response normalization, real time monitoring and feedback to response speed, etc..
The design of core process in this way can greatly improve the service quality of traditional two stages process, guarantee simultaneously
Very high efficiency of service.Because in the first stage, the advantage of machine is that contact staff is helped to solve simple question, to save
Less manpower cost, efficiency of service are also relatively high.But under traditional mode, the impression for ignoring client and mood body are tended to
It tests, this is that machine is less susceptible to know otherwise, and the exactly strong point of people.Therefore, second stage drawing as transition stage
Enter, can allow and manually carry out auxiliary machinery and judged, while provide buffer time, the body of such client to be transformed into manual service
Testing will be smoother, eliminates two stage " partition " or " tomography " sense.
The high efficiency smart customer service system of the present invention of embodiment 2
Fig. 2 is referred to, Fig. 2 is the structural block diagram of high efficiency smart customer service system of the present invention.The high efficiency smart customer service system
Include:
Client terminal 1: can be PC, plate or mobile phone.
Client's dialog interface module 2: i.e. the interface of terminal client and customer service system interaction, it is logical for realizing instant message
Letter, the proposition of problem and the presentation of response result are completed in client's dialog interface module 2.
Context of dialogue management module 3: it is responsible for the core business logic and contextual information of management of dialogs system.In dialogue
Hereafter the most important function of management module 3 is exactly to control the propulsion and circulation of three phase flows.Core business logic is:
1) context of dialogue management module 3 is responsible for receiving the chat messages of client's dialog interface module 2 in the first stage, adjusts
The enquirement of client is analyzed with intention identification module 7 and mood detection module 6, then search knowledge base 5 obtains crucial letter
Response template generation answer statement is ceased and utilized, answer statement is replied into client finally by client's dialog interface module 2.
2) if client can terminate to seek advice from simple two into three-wheel dialogue, entire to seek advice from answering knot
Beam;Otherwise context of dialogue management module 3 detects customer anger exception or intention assessment unstable result, then entering
Two-stage, at this time just start artificial customer service background module 4, send contact staff one end for response message, into customer service people
The supervision stage of member.
If 3) can complete to talk in second stage, entire answering of seeking advice from terminates;Otherwise contact staff's meeting
Enter the phase III according to field conditions selection, i.e., artificial customer service is leading, the stage of machine auxiliary.In this stage, in dialogue
Knowledge required for hereafter management module 3 can call knowledge base 5 that contact staff is helped to inquire, which is used as, to be assisted, such as commodity details,
Order information, logistics distribution information, counseling services specification masks of brand, etc..Meanwhile context of dialogue management module 3
Mood detection module 6 and intention assessment module 7 can be called to analyze the conversation content on both sides.
Artificial customer service background module 4: i.e. contact staff realizes the module that supervision and response are seeked advice from.
Knowledge base 5: it is used for the core data of response by RDF mode tissue, real-time query, institute are carried out using SparQL
State core data mainly and include the counseling services specification masks, etc. of commodity details, order information, logistics distribution information, brand.
Knowledge base 5 can constantly be expanded.
Mood detection module 6: whether normal for detecting mood during client questions.This is an independent technology
Module exports the emotional state that the sentence is reflected using the text of the consulting sentence of client as input.The master of mood detection
Stream method is namely based on detection of deep neural network, such as RNN classifier, LSTM classifier etc..
Intention assessment module 7: for detecting the intention of client questions.This is an independent technology modules, with client's
The text of sentence is seeked advice from as input, exports the intention of client expressed by the sentence.The main stream approach of intention assessment is namely based on
The detection of deep neural network, such as CNN classifier, Bi-LSTM+CRF parameter extraction method etc..
Quality and efficiency monitoring module 8: believed by acquiring the crucial of each round of links in dialog procedure in real time
Breath, efficiency and quality to dialogue carry out real-time quantization calculating and assessment.The data of acquisition include coming from client's dialog interface mould
The timestamp of the sentence of block 2, intention testing result and mood testing result from context of dialogue management module 3 come from people
The timestamp of the consulting response of work customer service background module 4.By the calculating of timestamp, available reply efficiency passes through intention
Detection and mood detect available customer satisfaction, thus the comprehensive service level assessment for obtaining response consulting.
Further, the quality and 8 comprehensive assessment of efficiency monitoring module go out service quality and efficiency of service.The service
The evaluation criteria of quality includes three factors: first is exactly the answer whether client is satisfied with, and second is exactly that client is connecing
During the entire process of being serviced whether be it is pleasant, third is exactly whether response content meets brand or the prepared service of businessman
Specification.The assessment of the efficiency of service is including the machine and artificial for entire service process, evaluation criteria include two because
Element: the round and response speed of dialogue.
Contact staff's terminal 9: can be PC, plate or mobile phone.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
Member, under the premise of not departing from the method for the present invention, can also make several improvement and supplement, these are improved and supplement also should be regarded as
Protection scope of the present invention.
Claims (10)
1. a kind of high efficiency smart client service method of Electronic Commerce field customer orientation, which comprises the following steps:
Step S1, in the first stage, context of dialogue management module are responsible for receiving the chat messages of client's dialog interface module, adjust
The enquirement of client is analyzed with intention identification module and mood detection module, then search knowledge base obtains key message simultaneously
Using response template generation answer statement, answer statement is replied into client finally by client's dialog interface module;
Step S2, if context of dialogue management module detects customer anger exception or intention assessment unstable result, then just
Into second stage, start artificial customer service background module, contact staff one end is sent by response message, into contact staff's
In the supervision stage, when the supervision stage for entering contact staff, the contextual information of dialogue can be presented on the auxiliary information at customer service interface
Region, to help contact staff to quickly understand the process entirely talked with;
Step S3, contact staff choose whether to enter the phase III according to the emotional state of existing customer, i.e., artificial customer service is leading,
The stage of machine auxiliary, in this stage, context of dialogue management module call knowledge base to help required for contact staff's inquiry
Knowledge as auxiliary;
Quality and efficiency monitoring module acquire the key message of each round of links in dialog procedure in real time, to Service Quality
Amount and efficiency of service carry out real-time quantization calculating and assessment;The evaluation criteria of service quality includes three factors: whether client obtains
Arrived satisfied answer, client during the entire process of receiving service whether be it is pleasant, response content whether meet brand or
The prepared service regulation of businessman;The assessment of efficiency of service is for entire service process, and evaluation criteria includes two factors:
The round and response speed of dialogue.
2. high efficiency smart client service method according to claim 1, which is characterized in that the quality and efficiency monitoring module are adopted
The key message of collection includes: the timestamp of the sentence from client's dialog interface module, from context of dialogue management module
It is intended to testing result and mood testing result, the timestamp of the consulting response from artificial customer service background module.
3. high efficiency smart client service method according to claim 2, which is characterized in that response is calculated by timestamp
Efficiency detects to obtain customer satisfaction by being intended to detection and mood.
4. high efficiency smart client service method according to claim 1, which is characterized in that the intention assessment module is with client's
The text of sentence is seeked advice from as input, exports the intention of client expressed by the sentence, it is intended that knowing method for distinguishing is based on depth mind
Detection through network;The mood detection module exports the sentence and is reflected using the text of the consulting sentence of client as input
The method of emotional state out, mood detection is the detection based on deep neural network.
5. high efficiency smart client service method according to claim 1, which is characterized in that the quality and efficiency monitoring module are real
When assessment service quality and efficiency of service, and timely feedback to machine and artificial customer service to make corresponding adjustment.
6. a kind of high efficiency smart customer service system of Electronic Commerce field customer orientation characterized by comprising
Client terminal;
Client's dialog interface module: i.e. the interface of terminal client and customer service system interaction, for realizing instant messaging;
Context of dialogue management module: being responsible for the core business logic and contextual information of management of dialogs system, controlled for three stages
The propulsion and circulation of process, the core business logic is:
1) context of dialogue management module is responsible for receiving the chat messages of client's dialog interface module in the first stage, calls and is intended to
Identification module and mood detection module analyze the enquirement of client, and then search knowledge base obtains key message and utilizes and answers
Template generation answer statement is answered, answer statement is replied into client finally by client's dialog interface module;
2) customer anger exception or intention assessment unstable result are detected such as context of dialogue management module, then entering the
Two-stage, start artificial customer service background module, send contact staff one end for response message, into the supervision rank of contact staff
Section, when the supervision stage for entering contact staff, the contextual information of dialogue can be presented on the auxiliary information region at customer service interface, with
Contact staff is helped to quickly understand the process entirely talked with;
3) contact staff chooses whether to enter the phase III according to the emotional state of existing customer, i.e., artificial customer service is leading, machine
The stage of auxiliary, in this stage, context of dialogue management module call knowledge base to help to know required for contact staff's inquiry
Know as auxiliary;
Artificial customer service background module: supervision and response consulting are realized for contact staff;
Knowledge base: tissue is used for the core data of response, carries out real-time query;
Mood detection module: whether normal for detecting mood during client questions;
Intention assessment module: for detecting the intention of client questions;
Quality and efficiency monitoring module: the key message by acquiring each round of links in dialog procedure in real time, it is right
Service quality and efficiency of service carry out real-time quantization calculating and assessment;The evaluation criteria of service quality includes three factors: client
The answer whether being satisfied with, client receive service during the entire process of whether be it is pleasant, whether response content meets
Brand or the prepared service regulation of businessman;The assessment of efficiency of service is for entire service process, and evaluation criteria includes two
A factor: the round and response speed of dialogue;
Contact staff's terminal.
7. high efficiency smart customer service system according to claim 6, which is characterized in that the quality and efficiency monitoring module are adopted
The key message of collection includes: the timestamp of the sentence from client's dialog interface module, from context of dialogue management module
It is intended to testing result and mood testing result, the timestamp of the consulting response from artificial customer service background module.
8. high efficiency smart customer service system according to claim 7, which is characterized in that response is calculated by timestamp
Efficiency detects to obtain customer satisfaction by being intended to detection and mood.
9. high efficiency smart customer service system according to claim 6, which is characterized in that the intention assessment module is with client's
The text of sentence is seeked advice from as input, exports the intention of client expressed by the sentence, it is intended that knowing method for distinguishing is based on depth mind
Detection through network;The mood detection module exports the sentence and is reflected using the text of the consulting sentence of client as input
The method of emotional state out, mood detection is the detection based on deep neural network.
10. high efficiency smart customer service system according to claim 6, which is characterized in that the quality and efficiency monitoring module
Real-time evaluation services quality and efficiency of service, and timely feedback to machine and artificial customer service to make corresponding adjustment.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910146974.3A CN109919671A (en) | 2019-02-27 | 2019-02-27 | The high efficiency smart client service method and system of a kind of Electronic Commerce field customer orientation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910146974.3A CN109919671A (en) | 2019-02-27 | 2019-02-27 | The high efficiency smart client service method and system of a kind of Electronic Commerce field customer orientation |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109919671A true CN109919671A (en) | 2019-06-21 |
Family
ID=66962691
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910146974.3A Pending CN109919671A (en) | 2019-02-27 | 2019-02-27 | The high efficiency smart client service method and system of a kind of Electronic Commerce field customer orientation |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109919671A (en) |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110533343A (en) * | 2019-09-04 | 2019-12-03 | 腾讯科技(深圳)有限公司 | Data processing method, device and the electronic equipment of intelligent customer service system |
CN110956472A (en) * | 2019-11-06 | 2020-04-03 | 深圳市一号互联科技有限公司 | Application method of man-machine coupling intelligent customer service platform |
CN111340585A (en) * | 2020-02-20 | 2020-06-26 | 上海索倍信息科技有限公司 | 3D panorama shopping e-commerce system based on block chain |
CN111885272A (en) * | 2020-07-24 | 2020-11-03 | 南京易米云通网络科技有限公司 | Intelligent call-out method for supporting telephone by call center seat and intelligent call center system |
CN112035325A (en) * | 2020-09-01 | 2020-12-04 | 中国银行股份有限公司 | Automatic monitoring method and device for text robot |
CN112699233A (en) * | 2019-10-17 | 2021-04-23 | ***通信集团浙江有限公司 | Service processing method and device and electronic equipment |
CN113079263A (en) * | 2021-03-16 | 2021-07-06 | 京东数字科技控股股份有限公司 | Method, device, system and medium for intelligent customer service switching |
CN113362084A (en) * | 2021-06-09 | 2021-09-07 | 广州智会云科技发展有限公司 | Client data tracking method, device, equipment and computer readable storage medium |
CN113704432A (en) * | 2021-08-31 | 2021-11-26 | 广州方舟信息科技有限公司 | Artificial intelligence customer service system construction method and device based on Internet hospital |
CN115221306A (en) * | 2022-09-21 | 2022-10-21 | 北京云迹科技股份有限公司 | Automatic response evaluation method and device |
CN115396389A (en) * | 2022-07-07 | 2022-11-25 | 深圳市华汇数据服务有限公司 | Internet of things information technology customer service system based on emotion energy perception |
CN115545799A (en) * | 2022-11-04 | 2022-12-30 | 北京赛西科技发展有限责任公司 | Information technology service quality evaluation method, device, equipment and medium |
CN115660465A (en) * | 2022-10-17 | 2023-01-31 | 杭州鸿骊科技有限公司 | Intelligent building comprehensive management system and method |
CN116303982A (en) * | 2023-05-24 | 2023-06-23 | 杭州实在智能科技有限公司 | Intelligent response and service processing method and system based on RPA and self-learning mechanism |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104809197A (en) * | 2015-04-24 | 2015-07-29 | 同程网络科技股份有限公司 | On-line question and answer method based on intelligent robot |
CN105591882A (en) * | 2015-12-10 | 2016-05-18 | 北京中科汇联科技股份有限公司 | Method and system for mixed customer services of intelligent robots and human beings |
CN108073976A (en) * | 2016-11-18 | 2018-05-25 | 科沃斯商用机器人有限公司 | Man-machine interactive system and its man-machine interaction method |
-
2019
- 2019-02-27 CN CN201910146974.3A patent/CN109919671A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104809197A (en) * | 2015-04-24 | 2015-07-29 | 同程网络科技股份有限公司 | On-line question and answer method based on intelligent robot |
CN105591882A (en) * | 2015-12-10 | 2016-05-18 | 北京中科汇联科技股份有限公司 | Method and system for mixed customer services of intelligent robots and human beings |
CN108073976A (en) * | 2016-11-18 | 2018-05-25 | 科沃斯商用机器人有限公司 | Man-machine interactive system and its man-machine interaction method |
Cited By (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110533343A (en) * | 2019-09-04 | 2019-12-03 | 腾讯科技(深圳)有限公司 | Data processing method, device and the electronic equipment of intelligent customer service system |
CN110533343B (en) * | 2019-09-04 | 2023-12-15 | 腾讯科技(深圳)有限公司 | Data processing method and device of intelligent customer service system and electronic equipment |
CN112699233A (en) * | 2019-10-17 | 2021-04-23 | ***通信集团浙江有限公司 | Service processing method and device and electronic equipment |
CN110956472A (en) * | 2019-11-06 | 2020-04-03 | 深圳市一号互联科技有限公司 | Application method of man-machine coupling intelligent customer service platform |
CN111340585A (en) * | 2020-02-20 | 2020-06-26 | 上海索倍信息科技有限公司 | 3D panorama shopping e-commerce system based on block chain |
CN111885272A (en) * | 2020-07-24 | 2020-11-03 | 南京易米云通网络科技有限公司 | Intelligent call-out method for supporting telephone by call center seat and intelligent call center system |
CN111885272B (en) * | 2020-07-24 | 2021-11-16 | 南京易米云通网络科技有限公司 | Intelligent call-out method for supporting telephone by call center seat and intelligent call center system |
CN112035325A (en) * | 2020-09-01 | 2020-12-04 | 中国银行股份有限公司 | Automatic monitoring method and device for text robot |
CN112035325B (en) * | 2020-09-01 | 2023-08-18 | 中国银行股份有限公司 | Text robot automatic monitoring method and device |
CN113079263A (en) * | 2021-03-16 | 2021-07-06 | 京东数字科技控股股份有限公司 | Method, device, system and medium for intelligent customer service switching |
CN113362084A (en) * | 2021-06-09 | 2021-09-07 | 广州智会云科技发展有限公司 | Client data tracking method, device, equipment and computer readable storage medium |
CN113704432A (en) * | 2021-08-31 | 2021-11-26 | 广州方舟信息科技有限公司 | Artificial intelligence customer service system construction method and device based on Internet hospital |
CN115396389A (en) * | 2022-07-07 | 2022-11-25 | 深圳市华汇数据服务有限公司 | Internet of things information technology customer service system based on emotion energy perception |
CN115396389B (en) * | 2022-07-07 | 2023-09-22 | 深圳市华汇数据服务有限公司 | Emotion energy perception-based Internet of things information technology customer service system |
CN115221306A (en) * | 2022-09-21 | 2022-10-21 | 北京云迹科技股份有限公司 | Automatic response evaluation method and device |
CN115660465A (en) * | 2022-10-17 | 2023-01-31 | 杭州鸿骊科技有限公司 | Intelligent building comprehensive management system and method |
CN115545799B (en) * | 2022-11-04 | 2023-03-24 | 北京赛西科技发展有限责任公司 | Information technology service quality evaluation method, device, equipment and medium |
CN115545799A (en) * | 2022-11-04 | 2022-12-30 | 北京赛西科技发展有限责任公司 | Information technology service quality evaluation method, device, equipment and medium |
CN116303982A (en) * | 2023-05-24 | 2023-06-23 | 杭州实在智能科技有限公司 | Intelligent response and service processing method and system based on RPA and self-learning mechanism |
CN116303982B (en) * | 2023-05-24 | 2023-08-22 | 杭州实在智能科技有限公司 | Intelligent response and service processing method and system based on RPA and self-learning mechanism |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109919671A (en) | The high efficiency smart client service method and system of a kind of Electronic Commerce field customer orientation | |
CN107329967B (en) | Question answering system and method based on deep learning | |
CN109710772B (en) | Question-answer base knowledge management system based on deep learning and implementation method thereof | |
CN110769124B (en) | Electric power marketing customer communication system | |
CN108170835A (en) | A kind of intelligent customer service system of combinatorial artificial and AI | |
CN104598445B (en) | Automatically request-answering system and method | |
CN104809197A (en) | On-line question and answer method based on intelligent robot | |
CN110263144A (en) | A kind of answer acquisition methods and device | |
CN107506372A (en) | A kind of automatic conversation switching method of robot customer service under mixed type session | |
CN109672794A (en) | A kind of outer paging system of intelligent sound | |
CN106951468A (en) | Talk with generation method and device | |
CN107562856A (en) | A kind of self-service customer service system and method | |
WO2016201815A1 (en) | Method and apparatus for providing online customer service | |
CN105261362A (en) | Conversation voice monitoring method and system | |
CN109446304A (en) | Intelligent customer service session method and system | |
CN111192060A (en) | Electric power IT service-based full-channel self-service response implementation method | |
CN109086371A (en) | A kind of semantic net interactive system and exchange method for legal advice | |
US8781880B2 (en) | System, method and apparatus for voice analytics of recorded audio | |
CN103458056A (en) | Speech intention judging method based on automatic classification technology for automatic outbound system | |
CN110188163A (en) | Data intelligence processing system based on natural language | |
CN113282737B (en) | Man-machine cooperation intelligent customer service dialogue method and device | |
CN112885348A (en) | AI-combined intelligent voice electric marketing method | |
CN111797202B (en) | Customer service conversation method based on artificial intelligence | |
KR20210125399A (en) | Interactive voice bot server and unmanned counsel system | |
TW202036323A (en) | Intelligent online customer service convergence core system which can provide a reply or query to an intention generated through voice recognition and natural language processing procedures |
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: 20190621 |