CN105808721A - Data mining based customer service content analysis method and system - Google Patents
Data mining based customer service content analysis method and system Download PDFInfo
- Publication number
- CN105808721A CN105808721A CN201610128454.6A CN201610128454A CN105808721A CN 105808721 A CN105808721 A CN 105808721A CN 201610128454 A CN201610128454 A CN 201610128454A CN 105808721 A CN105808721 A CN 105808721A
- Authority
- CN
- China
- Prior art keywords
- customer service
- client
- conversation content
- call
- described client
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
- G06F16/285—Clustering or classification
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Fuzzy Systems (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Mathematical Physics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a data mining based customer service content analysis method. The method comprises the steps of performing sentiment analysis on the dialogue content of a customer service staff and a customer to obtain sentiment values of the customer service staff and the customer; establishing a knowledge base and obtaining a call type and a call purpose of a dialogue between the customer service staff and the customer, wherein the call purpose comprises two composition attributes determined according to the call type that the dialogue content belongs to; combining the sentiment value of the customer service staff, the sentiment value of the customer, the call type between the customer service staff and the customer, and the two composition attributes corresponding to the call purpose to establish a five-element set of the dialogue content between the customer service staff and the customer; and analyzing the dialogue content between the customer service staff and the customer by utilizing the five-element set. According to the method, the five-element set of the dialogue content between the customer service staff and the customer is established for analyzing the dialogue content between the customer service staff and the customer to evaluate the service quality of the customer service staff.
Description
Technical field
The present invention relates to call center's customer service realm, particularly relate to a kind of customer service content analysis method based on data mining and system thereof.
Background technology
Along with enterprise and mechanism are to customer service Requirement Increases, customer service data are growing, and the content mining of customer service data is seemed extremely important.The content analysis of customer service data be it will be seen that the topic paid close attention to by customers in nearly a period of time, and the suggestion excavating client suitably adjusts business structure.The content analysis of customer service data precise positioning customer demand can be carried out individual business recommendation, set up and cultivate loyal client, for enterprise getting profit and establish good corporate image.The content analysis of each customer service data can be excavated the purpose of caller client in dialogue each time, whether be solved, so accurate analysis is possible not only to more fine granularity understand customer demand and can also obtain client's subjectivity and objective evaluation to customer service and conveniently improve customer service Quality advance CSAT.
Existing customer service QA system as emotional semantic classification task, uses disaggregated model to classify from dialogue it after extracting feature.Although such way can react the attitude to this section of dialogue of client effectively, part replaces client to give a mark realizing detection to customer service quality, but do not depict customer satisfaction or unsatisfied essence and reason, additionally also without the service quality evaluating customer service from objective angle, as client self engages in the dialogue with negative emotions.
At present, the knowledge base of conversation content needs professional and language specialist in field to lay down a regulation in a large number and consumes mental going to formulate, and do so is obviously incompatible with in advanced information society.In addition, owing to the content analysis in customer service field is based on the text after identifying, some wherein cannot have been avoided to identify mistake or oral expression, only gone throughout all of saying difficult especially by manpower.
Therefore, build intelligent customer service and analyze system, reach the high-level excavation to customer service content, thus constitutionally portrays the call intent of client more, just seem extremely urgent.
Summary of the invention
The present invention provides a kind of customer service content analysis method based on data mining and system, the method is based on the mode of semi-automatic construction of knowledge base and the conversation content between customer service and client is analyzed, it is not only able to detect the attitude of customer service, the purpose that namely theme of client and customer service dialogue calls can also be excavated, in addition can also detect whether customer service uses standard term of courtesy, if communicate with client with negative emotions.
The present invention the first invention provides a kind of customer service content analysis method based on data mining, and described method includes: the conversation content of customer service and client is carried out sentiment analysis, obtains the emotion value of described customer service and described client；Build knowledge base, obtain the type of call between described customer service with described client and call intent corresponding two constitutes attributes；Constitute combinations of attributes by corresponding to type of call between the emotion value of described customer service, the emotion value of described client, described customer service and described client and described call intent two, set up the five-tuple of conversation content between described customer service and described client；Utilize described five-tuple that conversation content between described customer service and described client is analyzed.
Preferably, the described conversation content to customer service and client carries out sentiment analysis, and the emotion value step obtaining described customer service and described client includes: based on the conversation content of described customer service Yu described client, each conversation content carries out two neutral or negative classification；And calculate the emotion value of each conversation content.
Preferably, described build knowledge base, obtain the type of call between described customer service with described client and call intent corresponding two constitutes attribute steps and includes: utilize data to formulate, build seed knowledge base, and then obtain the type of call between described customer service with described client and corresponding two of call intent constitute attributes.
Preferably, by the method learning word vector based on the degree of depth, seed knowledge base is expanded automatically.
Preferably, the emotion value of described client and described customer service obtains, and comprises the following steps: the conversation content between described client and described customer service is carried out pretreatment；According to the conventional emotion word of dialogue between described client and described customer service, carry out sentiment dictionary structure；Conversation content between described client and described customer service is carried out feature extraction, and feature includes sentiment dictionary, a gram language model unigram and two gram language model bigram；By model training, set up the sentiment classification model between described client and described customer service；The emotional category of each conversation content is predicted by described sentiment classification model；Integrate the emotional category of described each conversation content, obtain the emotion value of described client and described customer service.
Second aspect present invention provides a kind of customer service content analysis system based on data mining, and described system is between customer service and client, and it includes sentiment analysis module, call intent extraction module and conversation content analysis module；Wherein, described sentiment analysis module, by the conversation content between described customer service and described client is carried out sentiment analysis, obtain the emotion value of described customer service and described client；Described call intent extraction module, extracts the call intent of conversation content between described customer service and described client, it is thus achieved that two attributes that between described client with described customer service, the type of call of conversation content is corresponding with call intent；Described conversation content analysis module, emotion value according to described customer service, the emotion value of described client, the five-tuple of type of call and two attributes compositions corresponding to described call intent between described customer service and described client, be analyzed the conversation content of described customer service Yu described client.
Preferably, described system also includes base module, and described base module includes seed repository unit and the automatic expansion unit of seed knowledge base；Wherein, described seed repository unit, formulated by data, including two attributes that the type of call of conversation content between described customer service with described client and call intent are corresponding；The described automatic expansion unit of seed knowledge base, described seed repository unit is expanded by method automatically that learn word vector based on the degree of depth.
Conversation content between described customer service and described client, by setting up the five-tuple of conversation content between customer service and client, is analyzed by the present invention, and then evaluates the service quality of customer service.
Accompanying drawing explanation
In order to be illustrated more clearly that the technical scheme of the embodiment of the present invention, below the accompanying drawing used required during embodiment is described is briefly described, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the premise not paying creative work, it is also possible to obtain other accompanying drawing according to these accompanying drawings.
The schematic flow sheet of a kind of customer service content analysis method based on data mining that Fig. 1 provides for the embodiment of the present invention；
Fig. 2 obtains the block schematic illustration of five-tuple for a kind of customer service content analysis method based on data mining that the embodiment of the present invention provides；
The block schematic illustration of a kind of customer service content analysis system based on data mining that Fig. 3 provides for the embodiment of the present invention.
Detailed description of the invention
For making the purpose of the embodiment of the present invention, technical scheme and advantage clearly, below in conjunction with the accompanying drawing in the embodiment of the present invention, technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is a part of embodiment of the present invention, rather than whole embodiments.
The present invention provides a kind of customer service content analysis method based on data mining and system, the method is based on the mode of semi-automatic construction of knowledge base and the conversation content between customer service and client is analyzed, it is not only able to the attitude detecting customer service to service, can also excavate customer service attitude for the purpose that namely calls of theme, in addition can also detect whether client uses standard term of courtesy, if communicate with client with negative emotions.
For Fig. 1, the embodiment of the present invention is illustrated below, the schematic flow sheet of a kind of data analysing method that Fig. 1 provides for the embodiment of the present invention.As it can be seen, the method comprising the steps of S101-S104:
Step S101: the conversation content of customer service and client is carried out sentiment analysis, obtains the emotion value of described customer service and described client；
It should be noted that the conversation content of customer service Yu client is analyzed, it is by content of text digging technology.Content of text digging technology refers mainly to excavate valuable information from text data, belongs to machine learning application on text, such as text classification, information extraction, summary etc..Text automatically can be divided into specific classification by Text Classification.These classifications can be the emotion kind etc. of art, expression.Information extraction technique can automatically extract the customizing messages in text, such as keyword extraction etc..
Step S102: build knowledge base, obtains the type of call between described customer service with described client and call intent corresponding two constitutes attributes；
It should be noted that knowledge base mainly forms through information integration technique construction according to field priori, set up the knowledge base with field characteristic and can be effectively improved the accuracy of text mining.
Step S103: constitute combinations of attributes by corresponding to type of call between the emotion value of described customer service, the emotion value of described client, described customer service and described client and described call intent two, set up the five-tuple of conversation content between described customer service and described client；
Step S104: utilize described five-tuple that conversation content between described customer service and described client is analyzed.
In the above-mentioned methods, the described conversation content to customer service and client carries out sentiment analysis, the emotion value step obtaining described customer service and described client includes: based on the conversation content of described customer service Yu described client, each conversation content carries out two neutral or negative classification；And calculate the emotion value of each conversation content.
In the above-mentioned methods, described build knowledge base, obtain the type of call between described customer service with described client and call intent corresponding two constitutes attribute steps and includes: utilize data to formulate, build seed knowledge base, and then obtain the type of call between described customer service with described client and corresponding two of call intent constitute attributes.
In the above-mentioned methods, by the method learning word vector based on the degree of depth, seed knowledge base is expanded automatically.
In the above-mentioned methods, the emotion value of described client and described customer service obtains, and comprises the following steps: the conversation content between described client and described customer service is carried out pretreatment；According to the conventional emotion word of dialogue between described client and described customer service, carry out sentiment dictionary structure；Conversation content between described client and described customer service is carried out feature extraction, and feature includes sentiment dictionary, a gram language model unigram and two gram language model bigram；By model training, set up the sentiment classification model between described client and described customer service；The emotional category of each conversation content is predicted by described sentiment classification model；Integrate the emotional category of described each conversation content, obtain the emotion value of described client and described customer service.
Below for Fig. 2, above-mentioned five-tuple is illustrated.Fig. 2 obtains the block schematic illustration of five-tuple for a kind of customer service content analysis method based on data mining that the embodiment of the present invention provides.
First, the conversation content between client and customer service is carried out sentiment analysis, obtain client's emotion value and customer service emotion value.It is to say, the attitude obtaining client and customer service whether specification term of courtesy can be analyzed or with negative emotions speech etc. based on the conversation content between client and customer service.
In embodiments of the present invention, it is based on Sentence-level content (each conversation content) and carries out two classification and neutral or negative, calculated dialogue emotion value by Sentence-level classification results simultaneously.
In conversation content between client and customer service, carry out the acquisition of emotion value, comprise the following steps:
1, conversation content is carried out pretreatment, remove and disable symbol；
2, sentiment dictionary builds, and builds according to the conventional emotion word of communication between client and customer service；
3, feature extraction, the emotion word feature of unigram (individual character), bigram (double word) is extracted based on word, eigenvalue is IF-IDF value, the eigenvalue of sentiment dictionary is Boolean (i.e. logical value true or false, " 0 " or " 1 "), and conversation content is expressed as boolean vector form；That is each dimension represents Feature Words, and its value indicates whether there is such feature；
4, model training；
It should be noted that, it is current main stream approach that conversation content is carried out emotional semantic classification by the method adopting machine learning classification, and its conventional model mainly includes traditional classifier such as naive Bayesian (NB), support vector machine (SVM) etc. and neutral net such as convolutional neural networks (CNN) method；This is not limited by the embodiment of the present invention, no matter adopting which type of model to portray emotional semantic classification system, sample mainly (machine learning, is generally divided into independent three part training set trainset, checking collection validationset and test set testset from training set by disaggregated model；Wherein, training set is used for setting up model) learn to represent that a class another characteristic carries out parameter learning.
5, predicting the emotional category of each conversation content: it is identical with the training stage that the data that prediction solves prepare needs, characteristic meaning is corresponding；Then model is used each conversation content to be predicted and exports the probit that result is corresponding with result；
6, arrange Sentence-level to predict the outcome, and set threshold calculations and go out the emotion value of the client in each conversation content or customer service；And the probit of the emotion value obtained and emotion value is integrated, for instance: set threshold value.Finally give the customer service of conversation content rank and the output of the emotion value of client.
Secondly, build knowledge base, obtain the type of call of the conversation content between above-mentioned customer service and client and corresponding two of call intent constitute attributes.
Knowledge base refers to, mainly forms through information integration technique construction according to field priori, sets up the knowledge base with field characteristic and can be effectively improved the accuracy of text mining.The method building knowledge base has many, here the main method adopted based on degree of depth study word vector.There is a large amount of colloquial style due to customer service in talking with, identify after the few phenomenon such as word, wrongly written character of multiword add to use and manually go to build and the workload in maintenance knowledge storehouse is huge, the method learning word vector based on the degree of depth can carry out automatically expanding effectively solving the problems referred to above from artificial constructed seed knowledge base.
It should be noted that the present invention learns the method for word vector based on the semi-automatic degree of depth carries out the structure of knowledge base of client's suggestion.First the method trains word vector according to a large amount of customer service language materials, recycles the artificial constructed kind subframe storehouse that constantly expands knowledge based on similarity and finally tends to its integrity.The structure of knowledge base has directiveness for effective and accurate extraction call intent.
Below the knowledge base between customer service and client specifically being built and illustrate, its structure is broadly divided into two stages:
One, the building of seed knowledge base；At this time, it may be necessary to rely on artificial and data formulation, the seed knowledge base after building can summarize classification and its composition attribute of call conversation from macroscopic view, and can clearly portray the purpose of customer call.As in embodiments of the present invention, the classification elder generation rough sort of client's conversation content and business handling or problem consulting.For business handling classification, automatically add two attributes and action attributes and object properties, action attributes such as " open-minded ", " cancellation " etc., object properties such as " international roaming ", " broadband " etc.；For problem Category of consulting, then acquiescence is containing object properties and question attributes, object properties substantially similar have common factor but incomplete same but all belong to same attribute such as " mobile phone " with above-mentioned, " signal " etc., question attributes be then the problem that contains of description object as: (mobile phone) " is not shown ", (signal) " bad " etc..
Two, seed knowledge base is expanded by method based on degree of depth study word vector automatically；Owing to knowledge base goes structure or maintenance to be very huge workloads by manpower, and the cost that data updating decision is safeguarded is high every time, add in the content of text after identification unavoidable containing many identifications wrong or colloquial vocabulary, manually accomplishing all throughout being an extremely difficult job.In order to solve the problems referred to above, the embodiment of the present invention adopts the method based on degree of depth study word vector automatically to expand according to building seed knowledge base.The method needs the training of a large amount of customer service language materials to belong to the word vector in this field, mainly by following step:
1, needing corpus is carried out participle before training, general dictionary is difficult to meet and not only has spoken language but also have the data of territoriality, so needing constantly to accumulate in practice the dictionary in this field；
2, stop words is removed；
3, training word vector；
4, similarity is utilized to carry out expanding acquisition knowledge base on the model of seed knowledge base, it is thus achieved that to tend to complete knowledge base.Such as: in action attributes, in language material, often find the vocabulary similar to action attributes seed, calculate " open-minded " this vocabulary arise that " payment ", " activation ", " ruing well " these there is the vocabulary of same action attributes；Calculate " international roaming " this vocabulary and arise that " calling transfer ", " roaming bag " etc. have the vocabulary of same object properties.
And then, it is judged that type of call belonging to conversation content between client and customer service: problem Category of consulting or business handling classification；And extract the question attributes of problem Category of consulting and object properties, or the object properties of business handling classification or action attributes.Extraction step particularly as follows:
1, the classification of conversation content between customer service and client is extracted, for instance: if it has been determined that the classification of this conversation content is business handling class, then system default has action attributes and object properties for this call intent；
2, in knowledge base in action attributes and object properties, select portray a pair combination of this section of conversation content.Main application message extraction technique in this step, information extraction main stream approach adopts Mutual information entropy, left and right comentropy and TextRank algorithm.To this it should be noted that the method for information extraction is not limited by the embodiment of the present invention.
It is, according to the data content extracted, a pair combination that corresponding Attributions selection similarity is the highest in knowledge base respectively forms the tlv triple expressing this conversation content.Such as: problem consulting class, question attributes, object properties；Or business handling class, object properties, action attributes；
3, in conjunction with the emotion value of the emotion value of customer service and client, it is thus achieved that five-tuple；Such as: customer service emotion value, client's emotion value, problem Category of consulting, question attributes and object properties, or customer service emotion value, client's emotion value, business handling class, object properties and action attributes.
Finally, utilize the five-tuple of said extracted, it is possible to Accurate Analysis customer demand and attitude, also can directly or indirectly evaluate customer service quality simultaneously.Client is existed to the conversation content of negative emotion value, on the one hand can self attitude of subjective analysis client, the form secondly by five-tuple can also more comprehensively and objective analysis service quality.Can also propose to service more targetedly for client's purpose open question, increase customer satisfaction degree and loyalty.
The tlv triple of said extracted is that the purpose to customer call is extracted, and belongs to information retrieval application in practical field in text mining；Main from one section of text, content of text is carried out purpose extraction and portrays the intention that in this section of content of text, client expresses.In embodiments of the present invention, text mining is mainly used in customer service field, and it can extract the call intent of this section of communication from one section of conversation content according to the knowledge base built.Such as: known customer call type belongs to business handling class, dialogue purpose can be automatically extracted according to the knowledge of the category in knowledge base, as opened international roaming.Customer call purpose being extracted, it is possible to effectively analyze demand and the behavior of user, understanding client provides quality services for client；Open question can also be provided simultaneously and service more targetedly.
The block schematic illustration of a kind of customer service content analysis system based on data mining that Fig. 3 provides for the embodiment of the present invention.As it is shown on figure 3, this system includes sentiment analysis module, call intent extraction module and conversation content analysis module.
Specifically, sentiment analysis module, by the conversation content between described customer service and described client is carried out sentiment analysis, obtain the emotion value of described customer service and described client；Call intent extraction module, extracts the call intent of conversation content between described customer service and described client, it is thus achieved that two attributes that between described client with described customer service, the type of call of conversation content is corresponding with call intent；Conversation content analysis module, emotion value according to described customer service, the emotion value of described client, the five-tuple of type of call and two attributes compositions corresponding to described call intent between described customer service and described client, be analyzed the conversation content of described customer service Yu described client.
In said system, this system also includes base module, and described base module includes seed repository unit and the automatic expansion unit of seed knowledge base；Wherein, described seed repository unit, formulated by data, including two attributes that the type of call of conversation content between described customer service with described client and call intent are corresponding；The described automatic expansion unit of seed knowledge base, described seed repository unit is expanded by method automatically that learn word vector based on the degree of depth.
Conversation content between customer service and client, by setting up the five-tuple of conversation content between customer service and client, is analyzed by the present invention；By data analysis and then the service quality evaluating customer service or client.
Above-described detailed description of the invention; the purpose of the present invention, technical scheme and beneficial effect have been further described; it is it should be understood that; the foregoing is only the specific embodiment of the present invention; the protection domain being not intended to limit the present invention; all within the spirit and principles in the present invention, any amendment of making, equivalent replacement, improvement etc., should be included within protection scope of the present invention.
Claims (7)
1. the customer service content analysis method based on data mining, it is characterised in that described method includes:
The conversation content of customer service and client is carried out sentiment analysis, obtains the emotion value of described customer service and described client；
Building knowledge base, obtain the type of call between described customer service and described client and call intent, wherein call intent includes two composition attributes that the type of call belonging to conversation content determines；
Constitute combinations of attributes by corresponding to type of call between the emotion value of described customer service, the emotion value of described client, described customer service and described client and described call intent two, set up the five-tuple of conversation content between described customer service and described client；
Utilize described five-tuple that conversation content between described customer service and described client is analyzed.
2. method according to claim 1, it is characterised in that the described conversation content to customer service and client carries out sentiment analysis, the emotion value step obtaining described customer service and described client includes:
Based on the conversation content of described customer service Yu described client, each conversation content is carried out two neutral or negative classification；And calculate the emotion value of each conversation content.
3. method according to claim 1, it is characterised in that described in build knowledge base, obtain the type of call between described customer service with described client and call intent corresponding two constitute attribute steps and includes:
Utilize data to formulate, build seed knowledge base, and then obtain the type of call between described customer service with described client and corresponding two of call intent constitute attributes.
4. method according to claim 3, it is characterised in that seed knowledge base is expanded automatically by the method learning word vector based on the degree of depth.
5. method according to claim 2, it is characterised in that the emotion value of described client and described customer service obtains, and comprises the following steps:
Conversation content between described client and described customer service is carried out pretreatment；
According to the conventional emotion word of dialogue between described client and described customer service, carry out sentiment dictionary structure；
Conversation content between described client and described customer service is carried out feature extraction, and feature includes sentiment dictionary, a gram language model unigram and two gram language model bigram；
By model training, set up the sentiment classification model between described client and described customer service；
The emotional category of each conversation content is predicted by described sentiment classification model；
Integrate the emotional category of described each conversation content, obtain the emotion value of described client and described customer service.
6. the customer service content analysis system based on data mining, it is characterised in that described system is between customer service and client, and it includes sentiment analysis module, call intent extraction module and conversation content analysis module；Wherein,
Described sentiment analysis module, by the conversation content between described customer service and described client is carried out sentiment analysis, obtains the emotion value of described customer service and described client；
Described call intent extraction module, extracts the call intent of conversation content between described customer service and described client, it is thus achieved that two attributes that between described client with described customer service, the type of call of conversation content is corresponding with call intent；
Described conversation content analysis module, emotion value according to described customer service, the emotion value of described client, the five-tuple of type of call and two attributes compositions corresponding to described call intent between described customer service and described client, be analyzed the conversation content of described customer service Yu described client.
7. system according to claim 6, it is characterised in that described system also includes base module, described base module includes seed repository unit and the automatic expansion unit of seed knowledge base；Wherein,
Described seed repository unit, is formulated by data, including two attributes that the type of call of conversation content between described customer service with described client and call intent are corresponding；
The described automatic expansion unit of seed knowledge base, described seed repository unit is expanded by method automatically that learn word vector based on the degree of depth.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610128454.6A CN105808721A (en) | 2016-03-07 | 2016-03-07 | Data mining based customer service content analysis method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610128454.6A CN105808721A (en) | 2016-03-07 | 2016-03-07 | Data mining based customer service content analysis method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105808721A true CN105808721A (en) | 2016-07-27 |
Family
ID=56466886
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610128454.6A Pending CN105808721A (en) | 2016-03-07 | 2016-03-07 | Data mining based customer service content analysis method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105808721A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106407333A (en) * | 2016-09-05 | 2017-02-15 | 北京百度网讯科技有限公司 | Artificial intelligence-based spoken language query identification method and apparatus |
| CN106469145A (en) * | 2016-09-30 | 2017-03-01 | 中科鼎富（北京）科技发展有限公司 | Text emotion analysis method and device |
| CN107452405A (en) * | 2017-08-16 | 2017-12-08 | 北京易真学思教育科技有限公司 | A kind of method and device that data evaluation is carried out according to voice content |
| CN107766560A (en) * | 2017-11-03 | 2018-03-06 | 广州杰赛科技股份有限公司 | The evaluation method and system of customer service flow |
| CN107832294A (en) * | 2017-11-06 | 2018-03-23 | 广州杰赛科技股份有限公司 | Customer service quality evaluating method and device |
| CN107895230A (en) * | 2017-11-06 | 2018-04-10 | 广州杰赛科技股份有限公司 | Customer service quality evaluating method and device |
| CN108764753A (en) * | 2018-06-06 | 2018-11-06 | 平安科技（深圳）有限公司 | Test method, apparatus, computer equipment and the storage medium of business personnel's ability |
| CN110472041A (en) * | 2019-07-01 | 2019-11-19 | 浙江工业大学 | A kind of file classification method towards the online quality inspection of customer service |
| CN110519816A (en) * | 2019-08-22 | 2019-11-29 | 普联技术有限公司 | A kind of radio roaming control method, device, storage medium and terminal device |
| CN110909166A (en) * | 2019-11-28 | 2020-03-24 | 贝壳技术有限公司 | Method, apparatus, medium, and electronic device for improving session quality |
| CN111046179A (en) * | 2019-12-03 | 2020-04-21 | 哈尔滨工程大学 | Text classification method for open network question in specific field |
| CN111563167A (en) * | 2020-07-15 | 2020-08-21 | 智者四海（北京）技术有限公司 | Text classification system and method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102456344A (en) * | 2010-10-22 | 2012-05-16 | 中国电信股份有限公司 | System and method for analyzing customer behavior characteristic based on speech recognition technique |
| CN103106211A (en) * | 2011-11-11 | 2013-05-15 | ***通信集团广东有限公司 | Emotion recognition method and emotion recognition device for customer consultation texts |
| CN103811009A (en) * | 2014-03-13 | 2014-05-21 | 华东理工大学 | Smart phone customer service system based on speech analysis |
-
2016
- 2016-03-07 CN CN201610128454.6A patent/CN105808721A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102456344A (en) * | 2010-10-22 | 2012-05-16 | 中国电信股份有限公司 | System and method for analyzing customer behavior characteristic based on speech recognition technique |
| CN103106211A (en) * | 2011-11-11 | 2013-05-15 | ***通信集团广东有限公司 | Emotion recognition method and emotion recognition device for customer consultation texts |
| CN103811009A (en) * | 2014-03-13 | 2014-05-21 | 华东理工大学 | Smart phone customer service system based on speech analysis |
Non-Patent Citations (1)
| Title |
|---|
| 陈京亮: "《现代物流理论与实务》", 31 December 2012, 北京：中国铁道出版社 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106407333A (en) * | 2016-09-05 | 2017-02-15 | 北京百度网讯科技有限公司 | Artificial intelligence-based spoken language query identification method and apparatus |
| CN106469145A (en) * | 2016-09-30 | 2017-03-01 | 中科鼎富（北京）科技发展有限公司 | Text emotion analysis method and device |
| CN107452405A (en) * | 2017-08-16 | 2017-12-08 | 北京易真学思教育科技有限公司 | A kind of method and device that data evaluation is carried out according to voice content |
| CN107766560A (en) * | 2017-11-03 | 2018-03-06 | 广州杰赛科技股份有限公司 | The evaluation method and system of customer service flow |
| CN107832294A (en) * | 2017-11-06 | 2018-03-23 | 广州杰赛科技股份有限公司 | Customer service quality evaluating method and device |
| CN107895230A (en) * | 2017-11-06 | 2018-04-10 | 广州杰赛科技股份有限公司 | Customer service quality evaluating method and device |
| CN108764753A (en) * | 2018-06-06 | 2018-11-06 | 平安科技（深圳）有限公司 | Test method, apparatus, computer equipment and the storage medium of business personnel's ability |
| CN110472041A (en) * | 2019-07-01 | 2019-11-19 | 浙江工业大学 | A kind of file classification method towards the online quality inspection of customer service |
| CN110519816A (en) * | 2019-08-22 | 2019-11-29 | 普联技术有限公司 | A kind of radio roaming control method, device, storage medium and terminal device |
| CN110519816B (en) * | 2019-08-22 | 2021-09-10 | 普联技术有限公司 | Wireless roaming control method, device, storage medium and terminal equipment |
| CN110909166A (en) * | 2019-11-28 | 2020-03-24 | 贝壳技术有限公司 | Method, apparatus, medium, and electronic device for improving session quality |
| CN110909166B (en) * | 2019-11-28 | 2021-07-16 | 贝壳找房(北京)科技有限公司 | Method, apparatus, medium, and electronic device for improving session quality |
| CN111046179A (en) * | 2019-12-03 | 2020-04-21 | 哈尔滨工程大学 | Text classification method for open network question in specific field |
| CN111046179B (en) * | 2019-12-03 | 2022-07-15 | 哈尔滨工程大学 | Text classification method for open network question in specific field |
| CN111563167A (en) * | 2020-07-15 | 2020-08-21 | 智者四海（北京）技术有限公司 | Text classification system and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105808721A (en) | Data mining based customer service content analysis method and system | |
| CN110675288B (en) | Intelligent auxiliary judgment method, device, computer equipment and storage medium | |
| CN107329967B (en) | Question answering system and method based on deep learning | |
| CN104102723B (en) | Search for content providing and search engine | |
| US20230162051A1 (en) | Method, device and apparatus for execution of automated machine learning process | |
| CN106357942A (en) | Intelligent response method and system based on context dialogue semantic recognition | |
| CN109376251A (en) | A kind of microblogging Chinese sentiment dictionary construction method based on term vector learning model | |
| CN104076944A (en) | Chat emoticon input method and device | |
| TWI650719B (en) | System and method for evaluating customer service quality from text content | |
| CN104933113A (en) | Expression input method and device based on semantic understanding | |
| CN111651996A (en) | Abstract generation method and device, electronic equipment and storage medium | |
| CN105912629A (en) | Intelligent question and answer method and device | |
| CN109710766B (en) | Complaint tendency analysis early warning method and device for work order data | |
| CN113468296A (en) | Model self-iteration type intelligent customer service quality inspection system and method capable of configuring business logic | |
| CN109726253B (en) | Talent map and talent portrait construction method, device, equipment and medium | |
| CN112966082A (en) | Audio quality inspection method, device, equipment and storage medium | |
| CN105912645A (en) | Intelligent question and answer method and apparatus | |
| CN107203265A (en) | Information interacting method and device | |
| CN109634935A (en) | Method of speech processing, storage medium and device | |
| Bockhorst et al. | Predicting self-reported customer satisfaction of interactions with a corporate call center | |
| CA3182191A1 (en) | Voice quality inspection method and device, computer equipment and storage medium | |
| US10068567B1 (en) | System, method, and computer program for automatic management of intent classification | |
| CN108228779B (en) | Score prediction method based on learning community conversation flow | |
| US20220392434A1 (en) | Reducing biases of generative language models | |
| CN112579730A (en) | High-expansibility multi-label text classification method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160727 |
|
| RJ01 | Rejection of invention patent application after publication |