CN108256009A - A kind of method for improving electric intelligent response robot and answering accuracy rate - Google Patents
A kind of method for improving electric intelligent response robot and answering accuracy rate Download PDFInfo
- Publication number
- CN108256009A CN108256009A CN201810005077.6A CN201810005077A CN108256009A CN 108256009 A CN108256009 A CN 108256009A CN 201810005077 A CN201810005077 A CN 201810005077A CN 108256009 A CN108256009 A CN 108256009A
- Authority
- CN
- China
- Prior art keywords
- knowledge
- client
- similarity
- keyword
- feedback
- 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.)
- Granted
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
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of methods for improving electric intelligent response robot and answering accuracy rate, include the following steps：Client puts question in electronic channel；After electric intelligent response robot receives customer issue, semantic understanding is carried out to customer issue, i.e., is segmented and screened acquisition keyword to customer issue；The problem of keyword of acquisition with the knowledge in the knowledge base of backstage is matched, calculates client's input and the similarity of knowledge in knowledge base；Knowledge is subjected to descending arrangement according to gained similarity is calculated, the knowledge feedback for selecting text similarity higher is to client；By knowledge feedback to client after, if client does not click at the appointed time, ask client feedback the reason of not clicking.It is of the invention fully with reference to client to puing question to the feedback of result, improve robot and answer a question accuracy rate, promote customer service experience.
Description
Technical field
The present invention relates to power customer service technology fields, and in particular to a kind of raising electric intelligent response robot is answered
The method of accuracy rate.
Background technology
The universal idea and life style for having changed people deeply of internet, customer service form start from demand
Type changes to interactive.In addition, in recent years, artificial intelligence technology is grown rapidly, in weiqi play chess, automatic Pilot, image
The multiple fields such as identification, speech recognition achieve breakthrough.
In order to adapt to customer personalized demand for services, state's net Jiangsu electric power has actively promoted wechat public platform, mobile phone
Multiple electronic channels such as APP, and intelligent response robot is developed, to cope with client's demand for services of 24 hours, how to improve
Electric intelligent response robot answer accuracy rate, promoted customer experience, become one it is important the problem of.
Invention content
To solve deficiency of the prior art, the present invention provides a kind of raising electric intelligent response robot and answers accuracy rate
Method, solve the problems, such as electric intelligent response robot answer accuracy rate it is not high.
In order to realize above-mentioned target, the present invention adopts the following technical scheme that：A kind of raising electric intelligent response robot returns
The method for answering accuracy rate, it is characterised in that：Include the following steps：
First, client puts question in electronic channel；
2nd, after electric intelligent response robot receives customer issue, semantic understanding is carried out to customer issue, i.e., client is asked
Topic is segmented and is screened acquisition keyword；
3rd, the problem of keyword of acquisition being matched with the knowledge in the knowledge base of backstage, calculating client's input is with knowing
Know the similarity of knowledge in library；
4th, knowledge is subjected to descending arrangement according to gained similarity is calculated, the knowledge feedback for selecting text similarity higher
To client；
5th, by knowledge feedback to client after, if client does not click at the appointed time, ask the original that does not click of client feedback
Cause.
The method that accuracy rate is answered by a kind of aforementioned raising electric intelligent response robot, it is characterized in that：The electronics canal
Road includes webpage, wechat, APP.
The method that accuracy rate is answered by a kind of aforementioned raising electric intelligent response robot, it is characterized in that：The semantic reason
Solving detailed process is：
(1) using artificial intelligence semantic analysis technology, customer issue text is segmented according to natural language；
(2) vocabulary is deactivated according to power industry, rejects high frequency nonsense words and low frequency nonsense words are segmented
As a result, obtain keyword.
The method that accuracy rate is answered by a kind of aforementioned raising electric intelligent response robot, it is characterized in that：The electric power row
The deactivated vocabulary of industry is established process and is included：
1) text message involved by the knowledge base of backstage is segmented, obtains knowledge base word segmentation result word set；Calculate word
Concentrate inverse document frequency IDF of each word in entire knowledge basek；
In formula (1), IDFkFor the inverse document frequency of word k, total amounts of the I for knowledge in knowledge base, DkFor in knowledge base
Knowledge item number containing word k；
2) by the word segmentation result word set of knowledge base, screening is compared with existing general deactivated vocabulary, is weeded out in word set
Remaining word is carried out descending arrangement by general stop words according to its inverse document frequency；
3) number in the top specified in above-mentioned descending arrangement word set is chosen, by sales service personnel to it into pedestrian
Work is screened, and thinking that meaningless selected ci poem goes out in power industry, itself and general deactivated vocabulary is merged, become electric power row
Industry deactivates vocabulary.
The method that accuracy rate is answered by a kind of preceding raising electric intelligent response robot, it is characterized in that：It is described general
It deactivates vocabulary and uses " Baidu deactivates vocabulary ".
The method that accuracy rate is answered by a kind of aforementioned raising electric intelligent response robot, it is characterized in that：The client is defeated
The problem of entering and the knowledge Text similarity computing method of serial number i in knowledge base are as follows；
In formula (2), I is the total amount of knowledge in knowledge base, gained keyword quantity, t after N client questions semantic understandingsiFor
The problem of client inputs and the similarity of the knowledge text of serial number i, TFniGo out in the knowledge text of serial number i for n-th of keyword
Existing number, γiCoefficient is chosen for knowledge, P uses at least one in keyword 1 to keyword N for client before current time
The number that keyword is inquired, PiKnowledge for the serial number i client before current time is used in keyword 1- keywords N
The number that at least one keyword is selected in the case of being inquired.
The method that accuracy rate is answered by a kind of aforementioned raising electric intelligent response robot, it is characterized in that：It is described by knowledge
Descending arrangement is carried out according to gained similarity is calculated, the knowledge feedback for selecting text similarity higher is to client, specific choice side
Method is as follows：Assuming that the ranking of the similarity of serial number 1-5 knowledge is first to the 5th, t1~t5Its similarity is represented respectively：
(1) if (t1-t5)/t1≤ 0.5, then by the similarity ranking knowledge feedback of first five to client；
(2) if (t1-t5)/t1>0.5, (t1-t4)/t1≤ 0.5, then the knowledge feedback taken the first four place similarity is to client；
(3) if (t1-t5)/t1>0.5, (t1-t4)/t1>0.5, (t1-t3)/t1≤ 0.5, then by similarity ranking first three
Knowledge feedback is to client；
(4) if (t1-t5)/t1>0.5, (t1-t4)/t1>0.5, (t1-t3)/t1>0.5, (t1-t2)/t1≤ 0.5, then by phase
Like the degree ranking knowledge feedback of the first two to client；
(5) if (t1-t5)/t1>0.5, (t1-t4)/t1>0.5, (t1-t3)/t1>0.5, (t1-t2)/t1>0.5, then only by phase
Like the knowledge feedback that ranks the first of degree to client.
The method that accuracy rate is answered by a kind of aforementioned raising electric intelligent response robot, it is characterized in that：The electric power intelligence
Energy response robot system is integrated with artificial customer service system, when that can not answer customer issue, switches to artificial customer service.
The advantageous effect that the present invention is reached：
After electric intelligent response robot of the present invention receives customer issue, customer issue is segmented and is screened, and will
Word segmentation result is matched with the knowledge in the knowledge base of backstage, the problem of calculating client's input and acquainted text in knowledge base
After this similarity, descending arrangement is carried out to knowledge according to similarity, first five is selected to feed back to client；Simultaneously to similarity ranking
First five knowledge carries out postsearch screening, by the similarity of the 2nd to the 5th article of knowledge of ranking respectively with the similarity of the 1st article of knowledge into
Row compares, and the knowledge that similarity differs by more than 50% will be removed, to improve the matched precision of knowledge.
The problem of calculating client's input in knowledge base acquainted text similarity when, introduce knowledge and choose and be
Number after client is selected in the knowledge of feedback, records selection result and chooses coefficient for correcting knowledge, by the anti-of client
The matching result for optimizing knowledge is presented, the answer accuracy rate of intelligent robot can be effectively improved.
It is of the invention fully with reference to client to puing question to the feedback of result, improve robot and answer a question accuracy rate, promote client
Experience improves electric intelligent response robot and answers accuracy rate, robot can be used to replace operator attendance, section in wide range
Human-saving resource, while solve manually existing for answer the problems such as disunity, deviation, intelligent response robot is accurately and fast, entirely
The counseling services of period improve electrical power services efficiency and service quality, improve customer satisfaction, improve good service water
It is flat, corporate image is improved, expands social effectiveness.
Description of the drawings
Fig. 1 is flow chart of the method for the present invention.
Specific embodiment
The invention will be further described below in conjunction with the accompanying drawings.Following embodiment is only used for clearly illustrating the present invention
Technical solution, and be not intended to limit the protection scope of the present invention and limit the scope of the invention.
As shown in Figure 1, a kind of method for improving electric intelligent response robot and answering accuracy rate, includes the following steps：
First, client puts question in electronic channel, and electronic channel includes webpage, wechat, APP.
2nd, after electric intelligent response robot receives customer issue, semantic understanding is carried out to customer issue, i.e., client is asked
Topic is segmented and is screened acquisition keyword, and semantic understanding detailed process is：
(1) using artificial intelligence semantic analysis technology, customer issue text is segmented according to natural language；
(2) vocabulary is deactivated according to power industry, rejects high frequency nonsense words and low frequency nonsense words are segmented
As a result, it is N number of to obtain keyword；The deactivated dictionary of wherein power industry is established process and is included：
1) text message involved by the knowledge base of backstage is segmented, calculates the inverse in entire knowledge base of each word
Text frequency index IDFk。
In formula (1), IDFkFor the inverse document frequency of word k, total amounts of the I for knowledge in knowledge base, DkFor in knowledge base
Knowledge item number containing word k；
2) by the word segmentation result word set of knowledge base, screening is compared with existing general deactivated vocabulary, is weeded out in word set
General stop words, general deactivated vocabulary, will be remaining using " Baidu deactivates vocabulary " of the stop words situation that can represent every profession and trade
Word according to its inverse document frequency carry out descending arrangement.
3) number in the top specified in above-mentioned descending arrangement word set is chosen, such as first 500, by sales service personnel
Artificial screening is carried out to it, thinking that meaningless selected ci poem goes out in power industry, itself and general deactivated vocabulary are merged,
Vocabulary is deactivated as power industry.
3rd, the keyword of acquisition is matched with the knowledge in the knowledge base of backstage, matching detailed process is to calculate client
The problem of input and the similarity of the knowledge text of serial number i in knowledge base, Text similarity computing method are as follows；
In formula (2), I is the total amount of knowledge in knowledge base, gained keyword quantity, t after N client questions semantic understandingsiFor
The problem of client inputs and the similarity of the knowledge text of serial number i, TFniGo out in the knowledge text of serial number i for n-th of keyword
Existing number, γiCoefficient is chosen for knowledge, P uses at least one in keyword 1 to keyword N for client before current time
The number that keyword is inquired, PiKnowledge for the serial number i client before current time is used in keyword 1- keywords N
The number that at least one keyword is selected in the case of being inquired.
4th, knowledge is subjected to descending arrangement according to gained similarity is calculated, the knowledge feedback for selecting text similarity higher
To client, selection method is as follows：It is assumed that the ranking of the similarity of serial number 1-5 knowledge is first to the 5th, t1~t5It represents respectively
Its similarity.
(1) if (t1-t5)/t1≤ 0.5, then by the similarity ranking knowledge feedback of first five to client；
(2) if (t1-t5)/t1>0.5, (t1-t4)/t1≤ 0.5, then the knowledge feedback taken the first four place similarity is to client；
(3) if (t1-t5)/t1>0.5, (t1-t4)/t1>0.5, (t1-t3)/t1≤ 0.5, then by similarity ranking first three
Knowledge feedback is to client；
(4) if (t1-t5)/t1>0.5, (t1-t4)/t1>0.5, (t1-t3)/t1>0.5, (t1-t2)/t1≤ 0.5, then by phase
Like the degree ranking knowledge feedback of the first two to client；
(5) if (t1-t5)/t1>0.5, (t1-t4)/t1>0.5, (t1-t3)/t1>0.5, (t1-t2)/t1>0.5, then only by phase
Like the knowledge feedback that ranks the first of degree to client；
5th, by knowledge feedback to client after, if client does not click within stipulated time such as 30s, ask client feedback do not click
The reason of.
Electric intelligent response robot system is integrated with artificial customer service system, when that can not answer customer issue, is switched to
Artificial customer service.
After electric intelligent response robot of the present invention receives customer issue, customer issue is segmented and is screened, and will
Word segmentation result is matched with the knowledge in the knowledge base of backstage, the problem of calculating client's input and acquainted text in knowledge base
After this similarity, descending arrangement is carried out to knowledge according to similarity, first five is selected to feed back to client；Simultaneously to similarity ranking
First five knowledge carries out postsearch screening, by the similarity of the 2nd to the 5th article of knowledge of ranking respectively with the similarity of the 1st article of knowledge into
Row compares, and the knowledge that similarity differs by more than 50% will be removed, to improve the matched precision of knowledge.
The problem of calculating client's input in knowledge base acquainted text similarity when, introduce knowledge and choose and be
Number after client is selected in the knowledge of feedback, records selection result and chooses coefficient for correcting knowledge, by the anti-of client
The matching result for optimizing knowledge is presented, the answer accuracy rate of intelligent robot can be effectively improved.
The method that accuracy rate is answered by a kind of raising electric intelligent response robot of the present invention fully combines client to puing question to
As a result feedback, by the higher knowledge feedback of Documents Similarity to client, constantly improve robot answers a question accuracy rate, is promoted
Customer service is experienced.
The above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, without departing from the technical principles of the invention, several improvement and deformation can also be made, these are improved and deformation
Also it should be regarded as protection scope of the present invention.
Claims (8)
- A kind of 1. method for improving electric intelligent response robot and answering accuracy rate, it is characterised in that：Include the following steps：(1) client puts question in electronic channel；(2) electric intelligent response robot receive customer issue after, to customer issue carry out semantic understanding, i.e., to customer issue into Row participle and screening obtain keyword；(3) the problem of keyword of acquisition with the knowledge in the knowledge base of backstage being matched, calculating client's input and knowledge base The similarity of knowledge；(4) knowledge is subjected to descending arrangement according to gained similarity is calculated, the knowledge feedback for selecting text similarity higher is to visitor Family；(5) by knowledge feedback to client after, if client does not click at the appointed time, ask client feedback the reason of not clicking.
- 2. a kind of method for improving electric intelligent response robot and answering accuracy rate according to claim 1, it is characterized in that： The electronic channel includes webpage, wechat, APP.
- 3. a kind of method for improving electric intelligent response robot and answering accuracy rate according to claim 1, it is characterized in that： The semantic understanding detailed process is：(1) using artificial intelligence semantic analysis technology, customer issue text is segmented according to natural language；(2) vocabulary is deactivated according to power industry, rejects high frequency nonsense words and low frequency nonsense words obtain word segmentation result, Obtain keyword.
- 4. a kind of method for improving electric intelligent response robot and answering accuracy rate according to claim 3, it is characterized in that： The deactivated vocabulary of the power industry is established process and is included：(1) text message involved by the knowledge base of backstage is segmented, obtains knowledge base word segmentation result word set；It calculates in word set Inverse document frequency IDF of each word in entire knowledge basek；In formula (1), IDFkFor the inverse document frequency of word k, total amounts of the I for knowledge in knowledge base, DkTo contain in knowledge base The knowledge item number of word k；(2) by the word segmentation result word set of knowledge base, screening is compared with existing general deactivated vocabulary, is weeded out general in word set Remaining word is carried out descending arrangement by stop words according to its inverse document frequency；(3) number in the top specified in above-mentioned descending arrangement word set is chosen, it is carried out manually by sales service personnel Screening thinking that meaningless selected ci poem goes out in power industry, itself and general deactivated vocabulary is merged, become power industry Deactivate vocabulary.
- 5. a kind of method for improving electric intelligent response robot and answering accuracy rate according to claim 4, it is characterized in that： The general deactivated vocabulary uses " Baidu deactivates vocabulary ".
- 6. a kind of method for improving electric intelligent response robot and answering accuracy rate according to claim 1, it is characterized in that： The problem of client inputs is as follows with the knowledge Text similarity computing method of serial number i in knowledge base；In formula (2), I is the total amount of knowledge in knowledge base, gained keyword quantity, t after N client questions semantic understandingsiIt is defeated for client The problem of entering and the similarity of the knowledge text of serial number i, TFniTime occurred in the knowledge text of serial number i for n-th of keyword Number, γiCoefficient is chosen for knowledge, P uses at least one keyword in keyword 1 to keyword N for client before current time The number inquired, PiKnowledge for the serial number i client before current time uses at least 1 in keyword 1- keywords N The number that a keyword is selected in the case of being inquired.
- 7. a kind of method for improving electric intelligent response robot and answering accuracy rate according to claim 1, it is characterized in that： It is described that knowledge is subjected to descending arrangement according to gained similarity is calculated, the higher knowledge feedback of text similarity is selected to client, Specific choice method is as follows：Assuming that the ranking of the similarity of serial number 1-5 knowledge is first to the 5th, t1~t5Its phase is represented respectively Like degree：(1) if (t1-t5)/t1≤ 0.5, then by the similarity ranking knowledge feedback of first five to client；(2) if (t1-t5)/t1>0.5, (t1-t4)/t1≤ 0.5, then the knowledge feedback taken the first four place similarity is to client；(3) if (t1-t5)/t1>0.5, (t1-t4)/t1>0.5, (t1-t3)/t1≤ 0.5, then by the similarity ranking knowledge of first three Feed back to client；(4) if (t1-t5)/t1>0.5, (t1-t4)/t1>0.5, (t1-t3)/t1>0.5, (t1-t2)/t1≤ 0.5, then by similarity The ranking knowledge feedback of the first two is to client；(5) if (t1-t5)/t1>0.5, (t1-t4)/t1>0.5, (t1-t3)/t1>0.5, (t1-t2)/t1>0.5, then only by similarity The knowledge feedback to rank the first is to client.
- 8. a kind of method for improving electric intelligent response robot and answering accuracy rate according to claim 1, it is characterized in that： The electric intelligent response robot system is integrated with artificial customer service system, when that can not answer customer issue, is switched to artificial Customer service.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810005077.6A CN108256009B (en) | 2018-01-03 | 2018-01-03 | Method for improving answer accuracy of electric intelligent answer robot |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810005077.6A CN108256009B (en) | 2018-01-03 | 2018-01-03 | Method for improving answer accuracy of electric intelligent answer robot |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108256009A true CN108256009A (en) | 2018-07-06 |
| CN108256009B CN108256009B (en) | 2022-02-15 |
Family
ID=62725679
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810005077.6A Active CN108256009B (en) | 2018-01-03 | 2018-01-03 | Method for improving answer accuracy of electric intelligent answer robot |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108256009B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109446304A (en) * | 2018-10-10 | 2019-03-08 | 长沙师范学院 | Intelligent customer service session method and system |
| CN109727594A (en) * | 2018-12-27 | 2019-05-07 | 北京百佑科技有限公司 | Method of speech processing and device |
| CN109816449A (en) * | 2019-01-28 | 2019-05-28 | 广州供电局有限公司 | A kind of intelligent robot system for power marketing customer service |
| CN110222161A (en) * | 2019-05-07 | 2019-09-10 | 北京来也网络科技有限公司 | Talk with robot intelligent response method and device |
| CN110489518A (en) * | 2019-06-28 | 2019-11-22 | 北京捷通华声科技股份有限公司 | A kind of self-service feedback method and system based on feature extraction |
| CN110866089A (en) * | 2019-11-14 | 2020-03-06 | 国家电网有限公司 | Robot knowledge base construction system and method based on synonymous multi-language environment analysis |
| CN111104501A (en) * | 2019-12-23 | 2020-05-05 | 中国银行股份有限公司 | Knowledge determination method and device for test service |
| CN111475603A (en) * | 2019-01-23 | 2020-07-31 | 百度在线网络技术（北京）有限公司 | Enterprise identifier identification method and device, computer equipment and storage medium |
| CN112491649A (en) * | 2020-11-17 | 2021-03-12 | 中国平安财产保险股份有限公司 | Interface joint debugging test method and device, electronic equipment and storage medium |
| CN116955574A (en) * | 2023-09-19 | 2023-10-27 | 图林科技(深圳)有限公司 | Intelligent customer service robot based on artificial intelligence and application method thereof |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104657346A (en) * | 2015-01-15 | 2015-05-27 | 深圳市前海安测信息技术有限公司 | Question matching system and question matching system in intelligent interaction system |
| CN105843897A (en) * | 2016-03-23 | 2016-08-10 | 青岛海尔软件有限公司 | Vertical domain-oriented intelligent question and answer system |
| CN107153639A (en) * | 2016-03-04 | 2017-09-12 | 北大方正集团有限公司 | Intelligent answer method and system |
-
2018
- 2018-01-03 CN CN201810005077.6A patent/CN108256009B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104657346A (en) * | 2015-01-15 | 2015-05-27 | 深圳市前海安测信息技术有限公司 | Question matching system and question matching system in intelligent interaction system |
| CN107153639A (en) * | 2016-03-04 | 2017-09-12 | 北大方正集团有限公司 | Intelligent answer method and system |
| CN105843897A (en) * | 2016-03-23 | 2016-08-10 | 青岛海尔软件有限公司 | Vertical domain-oriented intelligent question and answer system |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109446304A (en) * | 2018-10-10 | 2019-03-08 | 长沙师范学院 | Intelligent customer service session method and system |
| CN109727594A (en) * | 2018-12-27 | 2019-05-07 | 北京百佑科技有限公司 | Method of speech processing and device |
| CN109727594B (en) * | 2018-12-27 | 2021-04-09 | 北京百佑科技有限公司 | Voice processing method and device |
| CN111475603A (en) * | 2019-01-23 | 2020-07-31 | 百度在线网络技术（北京）有限公司 | Enterprise identifier identification method and device, computer equipment and storage medium |
| CN109816449A (en) * | 2019-01-28 | 2019-05-28 | 广州供电局有限公司 | A kind of intelligent robot system for power marketing customer service |
| CN110222161A (en) * | 2019-05-07 | 2019-09-10 | 北京来也网络科技有限公司 | Talk with robot intelligent response method and device |
| CN110489518A (en) * | 2019-06-28 | 2019-11-22 | 北京捷通华声科技股份有限公司 | A kind of self-service feedback method and system based on feature extraction |
| CN110489518B (en) * | 2019-06-28 | 2021-09-17 | 北京捷通华声科技股份有限公司 | Self-service feedback method and system based on feature extraction |
| CN110866089A (en) * | 2019-11-14 | 2020-03-06 | 国家电网有限公司 | Robot knowledge base construction system and method based on synonymous multi-language environment analysis |
| CN110866089B (en) * | 2019-11-14 | 2023-04-28 | 国家电网有限公司 | Robot knowledge base construction system and method based on synonymous multi-context analysis |
| CN111104501A (en) * | 2019-12-23 | 2020-05-05 | 中国银行股份有限公司 | Knowledge determination method and device for test service |
| CN112491649A (en) * | 2020-11-17 | 2021-03-12 | 中国平安财产保险股份有限公司 | Interface joint debugging test method and device, electronic equipment and storage medium |
| CN116955574A (en) * | 2023-09-19 | 2023-10-27 | 图林科技(深圳)有限公司 | Intelligent customer service robot based on artificial intelligence and application method thereof |
| CN116955574B (en) * | 2023-09-19 | 2024-01-05 | 图林科技(深圳)有限公司 | Intelligent customer service robot based on artificial intelligence and application method thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108256009B (en) | 2022-02-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108256009A (en) | A kind of method for improving electric intelligent response robot and answering accuracy rate | |
| CN104504150B (en) | News public sentiment monitoring system | |
| CN107169001A (en) | A kind of textual classification model optimization method based on mass-rent feedback and Active Learning | |
| CN106095834A (en) | Intelligent dialogue method and system based on topic | |
| CN109871439B (en) | Question-answer community question routing method based on deep learning | |
| CN110033342A (en) | A kind of training method and device, a kind of recommended method and device of recommended models | |
| CN106572001B (en) | A kind of dialogue method and system of intelligent customer service | |
| CN110321421B (en) | Expert recommendation method for website knowledge community system and computer storage medium | |
| WO2016197855A1 (en) | Learning attainment recommendation method and device | |
| CN110347791B (en) | Topic recommendation method based on multi-label classification convolutional neural network | |
| CN106570181B (en) | Intelligent interaction method and system based on context management | |
| CN110992365A (en) | Loss function based on image semantic segmentation and design method thereof | |
| CN108765228A (en) | A kind of adaptive private teaching learning method of computer | |
| CN109558805A (en) | Human bodys' response method based on multilayer depth characteristic | |
| Azar et al. | Intelligent data mining and machine learning for mental health diagnosis using genetic algorithm | |
| CN111369290A (en) | Method and system for automatically generating data acquisition module | |
| CN111080087A (en) | Calling center scheduling method based on customer emotion analysis | |
| US20210081898A1 (en) | Human resource management system and method thereof | |
| CN112508108A (en) | Zero-sample Chinese character recognition method based on etymons | |
| CN116777148A (en) | Intelligent distribution processing system for service work orders based on data analysis | |
| CN108961453B (en) | Classroom roll call assistant decision-making system | |
| CN112308749B (en) | Culture plan generation device, method, electronic device, and readable storage medium | |
| CN107590156A (en) | A kind of polytypic method of text based on training set cyclic extension | |
| CN108845560A (en) | A kind of power scheduling log Fault Classification | |
| CN111064996B (en) | Method, system and storage medium for identifying user watching video content preference |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |