CN108256009B - Method for improving answer accuracy of electric intelligent answer robot - Google Patents
Method for improving answer accuracy of electric intelligent answer robot Download PDFInfo
- Publication number
- CN108256009B CN108256009B CN201810005077.6A CN201810005077A CN108256009B CN 108256009 B CN108256009 B CN 108256009B CN 201810005077 A CN201810005077 A CN 201810005077A CN 108256009 B CN108256009 B CN 108256009B
- Authority
- CN
- China
- Prior art keywords
- knowledge
- client
- similarity
- words
- customer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000012216 screening Methods 0.000 claims abstract description 17
- 230000011218 segmentation Effects 0.000 claims abstract description 16
- 238000004364 calculation method Methods 0.000 claims abstract description 8
- 238000005516 engineering process Methods 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 claims description 3
- 238000010187 selection method Methods 0.000 claims description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000009849 deactivation Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
Images
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 method for improving the answer accuracy of an electric intelligent answering robot, which comprises the following steps: the customer asks questions in the electronic channel; after receiving the customer problem, the electric intelligent response robot carries out semantic understanding on the customer problem, namely, carrying out word segmentation and screening on the customer problem to obtain a keyword; matching the obtained keywords with knowledge in a background knowledge base, and calculating the similarity between the problems input by the client and the knowledge in the knowledge base; the knowledge is sorted in descending order according to the similarity obtained by calculation, and the knowledge with higher text similarity is selected and fed back to the client; and after the knowledge is fed back to the client, if the client does not click within the specified time, requesting the client to feed back the reason of the non-click. The invention fully combines the feedback of the customer to the question result, improves the accuracy rate of answering the question by the robot and improves the customer service experience.
Description
Technical Field
The invention relates to the technical field of electric power customer service, in particular to a method for improving the answer accuracy of an electric power intelligent response robot.
Background
The popularity of the internet has profoundly changed people's conception and lifestyle, and the form of customer service has begun to shift from appealing to interactive. In addition, in recent years, the development of artificial intelligence technology has been rapidly advanced, and breakthrough progress has been made in a plurality of fields such as go playing, automatic driving, image recognition, voice recognition, and the like.
In order to meet the personalized service requirements of customers, the national network Jiangsu electric power actively popularizes a plurality of electronic channels such as WeChat public platforms and mobile phone APPs, and develops an intelligent response robot so as to meet the 24-hour service requirements of customers, how to improve the response accuracy of the electric intelligent response robot, and how to improve the customer experience, which becomes an important problem.
Disclosure of Invention
In order to solve the defects in the prior art, the invention provides a method for improving the response accuracy of an electric intelligent response robot, and solves the problem that the response accuracy of the electric intelligent response robot is not high.
In order to achieve the above purpose, the invention adopts the following technical scheme: a method for improving the answer accuracy of an electric intelligent answer robot is characterized by comprising the following steps: the method comprises the following steps:
firstly, a customer asks questions in an electronic channel;
after receiving the customer problems, the electric power intelligent response robot carries out semantic understanding on the customer problems, namely, carrying out word segmentation and screening on the customer problems to obtain keywords;
matching the obtained keywords with knowledge in a background knowledge base, and calculating the similarity between the problems input by the client and the knowledge in the knowledge base;
fourthly, sorting the knowledge in a descending order according to the similarity obtained by calculation, and selecting the knowledge with higher text similarity to feed back to the client;
and fifthly, after the knowledge is fed back to the client, if the client does not click within the specified time, the client is requested to feed back the reason of the non-click.
The method for improving the answer accuracy of the intelligent electric power answering robot is characterized in that: the electronic channel comprises a webpage, WeChat and APP.
The method for improving the answer accuracy of the intelligent electric power answering robot is characterized in that: the semantic understanding specific process comprises the following steps:
(1) segmenting the problem text of the client according to natural language by utilizing an artificial intelligent semantic analysis technology;
(2) and according to the word list disabled in the power industry, removing the high-frequency nonsense words and the low-frequency nonsense words to obtain word segmentation results, and obtaining keywords.
The method for improving the answer accuracy of the intelligent electric power answering robot is characterized in that: the establishment process of the electric power industry stop word list comprises the following steps:
1) and segmenting the text information related to the background knowledge base to obtain a knowledge base segmentation result word set.
Calculating the inverse text frequency index IDF of each word in the word set in the whole knowledge basek。
In the formula (1), IDFkIs the inverse text frequency index of the word k, I isTotal amount of knowledge in the knowledge base, DkThe number of knowledge items containing the word k in the knowledge base is shown;
2) and comparing and screening the word set of the word segmentation result of the knowledge base with the conventional universal stop word list, screening out the universal stop words in the word set, and sequencing the rest words in a descending order according to the inverse text frequency index of the rest words.
3) And selecting the number of words in the word set after descending order, wherein the rank of the words is appointed at the top, manually screening the words by marketing service personnel, selecting the words which are regarded as nonsense in the power industry, and combining the words with the conventional universal stop word list to form the stop word list in the power industry.
The method for improving the answer accuracy of the intelligent electric power answering robot is characterized in that: the conventional universal stop word list adopts a hundred-degree stop word list.
The method for improving the answer accuracy of the intelligent electric power answering robot is characterized in that: the similarity calculation method of the questions input by the client and the knowledge text with the sequence number i in the knowledge base is as follows;
in the formula (2), I is the total amount of knowledge in the knowledge base, INThe number of key words in the knowledge base is shown, N is the number of key words obtained after the client asks for semantic understanding, tiSimilarity of question input for client and knowledge text of sequence number i, TFniFor the number of times the nth keyword appears in the knowledge text of sequence number i, γiSelecting a coefficient for knowledge, wherein P is the number of times that a client uses at least 1 keyword from the keywords 1 to the keyword N to inquire before the current time, and PiThe number of times the knowledge of sequence number i is selected in case the client makes a query using at least 1 keyword of keywords 1 to N before the current time.
The method for improving the answer accuracy of the intelligent electric power answering robot is characterized in that: the knowledge is sorted in descending order according to the similarity obtained by calculation, and the knowledge with higher text similarity is selected and fed back to the client, wherein the specific selection method comprises the following steps: let rank names of similarity of knowledge of sequence numbers 1-5 be first to fifth, t1～t5Respectively, the similarity thereof:
(1) if (t)1-t5)/t1If the similarity is less than or equal to 0.5, feeding back the knowledge of the top five ranked similarities to the client;
(2) if (t)1-t5)/t1>0.5，(t1-t4)/t1If the similarity is less than or equal to 0.5, feeding back the knowledge of the first four of the ranking of the similarity to the client;
(3) if (t)1-t5)/t1>0.5，(t1-t4)/t1>0.5，(t1-t3)/t1If the similarity is less than or equal to 0.5, feeding back the knowledge of the top three ranked similarities to the client;
(4) if (t)1-t5)/t1>0.5，(t1-t4)/t1>0.5，(t1-t3)/t1>0.5，(t1-t2)/t1If the similarity is less than or equal to 0.5, feeding back the knowledge of the second highest ranking of the similarity to the client;
(5) if (t)1-t5)/t1>0.5，(t1-t4)/t1>0.5，(t1-t3)/t1>0.5，(t1-t2)/t1>0.5, only the knowledge of the first degree of similarity ranking is fed back to the client.
The method for improving the answer accuracy of the intelligent electric power answering robot is characterized in that: the electric power intelligent response robot system is integrated with the artificial customer service system, and when the customer questions cannot be answered, the electric power intelligent response robot system is switched to the artificial customer service system.
The invention achieves the following beneficial effects:
after receiving the customer problems, the electric intelligent response robot carries out word segmentation and screening on the customer problems, matches word segmentation results with knowledge in a background knowledge base, carries out descending order arrangement on the knowledge according to the similarity after calculating the similarity between the problems input by the customer and the known texts in the knowledge base, and feeds back five previous items to the customer; and simultaneously, carrying out secondary screening on the knowledge with the similarity ranking in the first five, comparing the similarity of the knowledge ranking from the 2 nd knowledge to the 5 th knowledge with the similarity of the knowledge ranking from the 1 st knowledge respectively, and rejecting the knowledge with the similarity difference exceeding 50% so as to improve the precision of knowledge matching.
When the similarity between the questions input by the client and the texts known in the knowledge base is calculated, the knowledge selection coefficient is introduced, after the client selects the feedback knowledge, the selection result is recorded and used for correcting the knowledge selection coefficient, the feedback of the client is used for optimizing the matching result of the knowledge, and the answer accuracy of the intelligent robot can be effectively improved.
The intelligent answering robot fully combines the feedback of the customer to the question result, improves the accuracy rate of answering the question by the robot, improves the customer experience, improves the accuracy rate of answering the question by the electric intelligent answering robot, can use the robot to replace a manual seat in a larger range, saves the human resource, solves the problems of non-uniformity, deviation and the like of manual answering, improves the electric power service efficiency and the service quality by the intelligent answering robot through accurate, quick and full-time consultation service, improves the customer satisfaction degree, improves the high-quality service level, improves the enterprise image and enlarges the social influence.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Detailed Description
The invention is further described below with reference to the accompanying drawings. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.
As shown in fig. 1, a method for improving the answer accuracy of an intelligent electric answering robot includes the following steps:
firstly, a customer asks questions in an electronic channel, wherein the electronic channel comprises a webpage, WeChat and APP.
After receiving the customer problem, the electric power intelligent response robot carries out semantic understanding on the customer problem, namely, the customer problem is subjected to word segmentation and screening to obtain a keyword, and the specific process of semantic understanding is as follows:
(1) segmenting the problem text of the client according to natural language by utilizing an artificial intelligent semantic analysis technology;
(2) according to a word list disabled in the power industry, high-frequency nonsense words and low-frequency nonsense words are removed to obtain word segmentation results, and N keywords are obtained; the establishment process of the electric power industry deactivation word bank comprises the following steps:
1) segmenting words of text information related to the background knowledge base, and calculating the inverse text frequency index IDF of each word in the whole knowledge basek。
In the formula (1), IDFkIs the inverse text frequency index of the word k, I is the total amount of knowledge in the knowledge base, DkThe number of knowledge items containing the word k in the knowledge base is shown;
2) and comparing and screening the word segmentation result word set of the knowledge base with the conventional universal stop word list, screening out the universal stop words in the word set, wherein the conventional universal stop word list adopts a Baidu stop word list capable of representing stop word conditions of various industries, and the rest words are arranged in a descending order according to the inverse text frequency index of the word.
3) Selecting the number of words in the descending order ordered word set, such as the first 500 words, which is designated by the front, manually screening the words by marketing service personnel, selecting words which are regarded as nonsense in the power industry, and combining the words with the conventional universal stop word list to form the stop word list in the power industry.
Matching the obtained keywords with knowledge in a background knowledge base, wherein the specific matching process is to calculate the similarity between the problems input by the client and the knowledge text with the sequence number i in the knowledge base, and the text similarity calculation method is as follows;
in the formula (2), I is the total amount of knowledge in the knowledge base, INThe number of key words in the knowledge base is shown, N is the number of key words obtained after the client asks for semantic understanding, tiSimilarity of question input for client and knowledge text of sequence number i, TFniFor the number of times the nth keyword appears in the knowledge text of sequence number i, γiSelecting a coefficient for knowledge, wherein P is the number of times that a client uses at least 1 keyword from the keywords 1 to the keyword N to inquire before the current time, and PiThe number of times the knowledge of sequence number i is selected in case the client makes a query using at least 1 keyword of keywords 1 to N before the current time.
Fourthly, sorting the knowledge in a descending order according to the similarity obtained by calculation, and selecting the knowledge with higher text similarity to feed back to the client, wherein the selection method comprises the following steps: assume that the rank names of the similarities of knowledge of sequence numbers 1-5 are first through fifth, t1～t5Respectively, the degrees of similarity thereof.
(1) If (t)1-t5)/t1If the similarity is less than or equal to 0.5, feeding back the knowledge of the top five ranked similarities to the client;
(2) if (t)1-t5)/t1>0.5，(t1-t4)/t1If the similarity is less than or equal to 0.5, feeding back the knowledge of the first four of the ranking of the similarity to the client;
(3) if (t)1-t5)/t1>0.5，(t1-t4)/t1>0.5，(t1-t3)/t1If the similarity is less than or equal to 0.5, feeding back the knowledge of the top three ranked similarities to the client;
(4) if (t)1-t5)/t1>0.5，(t1-t4)/t1>0.5，(t1-t3)/t1>0.5，(t1-t2)/t1If the similarity is less than or equal to 0.5, feeding back the knowledge of the second highest ranking of the similarity to the client;
(5) if (t)1-t5)/t1>0.5，(t1-t4)/t1>0.5，(t1-t3)/t1>0.5，(t1-t2)/t1>0.5, only the knowledge with the first similarity rank is fed back to the client;
and fifthly, after the knowledge is fed back to the client, if the client does not click within a specified time such as 30s, requesting the client to feed back the reason of the non-click.
The electric power intelligent response robot system is integrated with the artificial customer service system, and when the customer questions cannot be answered, the electric power intelligent response robot system is switched to the artificial customer service system.
After receiving the customer problems, the electric intelligent response robot carries out word segmentation and screening on the customer problems, matches word segmentation results with knowledge in a background knowledge base, carries out descending order arrangement on the knowledge according to the similarity after calculating the similarity between the problems input by the customer and the known texts in the knowledge base, and feeds back five previous items to the customer; and simultaneously, carrying out secondary screening on the knowledge with the similarity ranking in the first five, comparing the similarity of the knowledge ranking from the 2 nd knowledge to the 5 th knowledge with the similarity of the knowledge ranking from the 1 st knowledge respectively, and rejecting the knowledge with the similarity difference exceeding 50% so as to improve the precision of knowledge matching.
When the similarity between the questions input by the client and the texts known in the knowledge base is calculated, the knowledge selection coefficient is introduced, after the client selects the feedback knowledge, the selection result is recorded and used for correcting the knowledge selection coefficient, the feedback of the client is used for optimizing the matching result of the knowledge, and the answer accuracy of the intelligent robot can be effectively improved.
According to the method for improving the answer accuracy of the intelligent power response robot, feedback of a customer on a questioning result is fully combined, knowledge with high document similarity is fed back to the customer, the accuracy of answering questions by the robot is continuously improved, and customer service experience is improved.
The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.
Claims (5)
1. A method for improving the answer accuracy of an electric intelligent answer robot is characterized by comprising the following steps: the method comprises the following steps:
(1) the customer asks questions in the electronic channel;
(2) after receiving the customer problem, the electric intelligent response robot carries out semantic understanding on the customer problem, namely, carrying out word segmentation and screening on the customer problem to obtain a keyword;
(3) matching the obtained keywords with knowledge in a background knowledge base, and calculating the similarity between the problems input by the client and the knowledge in the knowledge base;
(4) the knowledge is sorted in descending order according to the similarity obtained by calculation, and the knowledge with higher text similarity is selected and fed back to the client;
(5) after the knowledge is fed back to the client, if the client does not click within the specified time, asking the client to feed back the reason for not clicking;
the knowledge is sorted in a descending order according to the similarity obtained by calculation, and the knowledge with higher text similarity is selected and fed back to the client, wherein the specific selection method comprises the following steps: let rank names of similarity of knowledge of sequence numbers 1-5 be first to fifth, t1～t5Respectively, the similarity thereof:
(1) if (t)1-t5)/t1If the similarity is less than or equal to 0.5, feeding back the knowledge of the top five ranked similarities to the client;
(2) if (t)1-t5)/t1>0.5，(t1-t4)/t1If the similarity is less than or equal to 0.5, feeding back the knowledge of the first four of the ranking of the similarity to the client;
(3) if (t)1-t5)/t1>0.5，(t1-t4)/t1>0.5，(t1-t3)/t1If the similarity is less than or equal to 0.5, feeding back the knowledge of the top three ranked similarities to the client;
(4) if (t)1-t5)/t1>0.5，(t1-t4)/t1>0.5，(t1-t3)/t1>0.5，(t1-t2)/t1If the similarity is less than or equal to 0.5, feeding back the knowledge of the second highest ranking of the similarity to the client;
(5) if (t)1-t5)/t1>0.5，(t1-t4)/t1>0.5，(t1-t3)/t1>0.5，(t1-t2)/t1>0.5, only the knowledge with the first similarity rank is fed back to the client;
the semantic understanding specific process comprises the following steps:
(1) segmenting the problem text of the client according to natural language by utilizing an artificial intelligent semantic analysis technology;
(2) according to a word list disabled in the power industry, high-frequency nonsense words and low-frequency nonsense words are removed to obtain word segmentation results, and keywords are obtained;
the establishment process of the electric power industry stop word list comprises the following steps:
(1) segmenting words of text information related to a background knowledge base to obtain a word set of word segmentation results of the knowledge base; calculating the inverse text frequency index IDF of each word in the word set in the whole knowledge basek：
In the formula (1), IDFkIs the inverse text frequency index of the word k, I is the total amount of knowledge in the knowledge base, DkThe number of knowledge items containing the word k in the knowledge base is shown;
(2) comparing and screening the word set of the word segmentation result of the knowledge base with the existing general stop word list, screening out general stop words in the word set, and arranging the rest words in a descending order according to the inverse text frequency index of the rest words;
(3) and selecting the number of words in the word set after descending order, wherein the rank of the words is appointed at the top, manually screening the words by marketing service personnel, selecting the words which are regarded as nonsense in the power industry, and combining the words with the conventional universal stop word list to form the stop word list in the power industry.
2. The method for improving the answer accuracy of the intelligent electric answering robot as claimed in claim 1, wherein: the electronic channel comprises a webpage and an APP.
3. The method for improving the answer accuracy of the intelligent electric answering robot as claimed in claim 1, wherein: the conventional universal stop word list adopts a hundred-degree stop word list.
4. The method for improving the answer accuracy of the intelligent electric answering robot as claimed in claim 1, wherein: the method for calculating the similarity between the question input by the client and the knowledge text with the sequence number i in the knowledge base comprises the following steps:
in the formula (2), I is the total amount of knowledge in the knowledge base, INThe number of key words in the knowledge base is shown, N is the number of key words obtained after the client asks for semantic understanding, tiSimilarity of question input for client and knowledge text of sequence number i, TFniFor the number of times the nth keyword appears in the knowledge text of sequence number i, γiSelecting a coefficient for knowledge, wherein P is the number of times that a client uses at least 1 keyword from the keywords 1 to the keyword N to inquire before the current time, and PiThe number of times the knowledge of sequence number i is selected in case the client makes a query using at least 1 keyword of keywords 1 to N before the current time.
5. The method for improving the answer accuracy of the intelligent electric answering robot as claimed in claim 1, wherein: the electric power intelligent response robot system is integrated with the artificial customer service system, and when the customer questions cannot be answered, the electric power intelligent response robot system is switched to the artificial customer service system.
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 CN108256009A (en) | 2018-07-06 |
| CN108256009B true 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) |
Families Citing this family (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 |
| CN109727594B (en) * | 2018-12-27 | 2021-04-09 | 北京百佑科技有限公司 | Voice processing method and device |
| CN111475603B (en) * | 2019-01-23 | 2023-07-04 | 百度在线网络技术（北京）有限公司 | Enterprise identification recognition method, enterprise identification recognition 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 |
| CN110222161B (en) * | 2019-05-07 | 2022-10-14 | 北京来也网络科技有限公司 | Intelligent response method and device for conversation robot |
| 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 |
| 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 |
| 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 |
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 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108256009A (en) | 2018-07-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108256009B (en) | Method for improving answer accuracy of electric intelligent answer robot | |
| CN110008399B (en) | Recommendation model training method and device, and recommendation method and device | |
| US20180157960A1 (en) | Scalable curation system | |
| US20230043174A1 (en) | Method for pushing anchor information, computer device, and storage medium | |
| CN109272129B (en) | Call center service management system | |
| US8566306B2 (en) | Scalable search system using human searchers | |
| US20140114986A1 (en) | Method and apparatus for implicit topic extraction used in an online consultation system | |
| CN110046230B (en) | Method for generating recommended speaking collection, and recommended speaking method and device | |
| CN106651544B (en) | Conversational recommendation system with minimal user interaction | |
| US20120296897A1 (en) | Text to Image Translation | |
| CN112507094B (en) | Customer service robot dialogue method based on reinforcement learning and related components thereof | |
| WO2008011537A2 (en) | Method, system, and computer readable medium useful in managing a computer-based system for servicing user initiated tasks | |
| CN108664515B (en) | A kind of searching method and device, electronic equipment | |
| CN108509461A (en) | A kind of sequence learning method and server based on intensified learning | |
| CN110516057B (en) | Petition question answering method and device | |
| CN110321421B (en) | Expert recommendation method for website knowledge community system and computer storage medium | |
| CN111080087B (en) | Calling center scheduling method based on customer emotion analysis | |
| TW201935370A (en) | System and method for evaluating customer service quality from text content | |
| CN102640146A (en) | Database searching method, system and controller | |
| CN113076423A (en) | Data processing method and device and data query method and device | |
| CN115292470B (en) | Semantic matching method and system for intelligent customer service of petty loan | |
| CN111881172A (en) | Question recommendation system based on answer statistical characteristics | |
| CN109033428B (en) | Intelligent customer service method and system | |
| CN104298789B (en) | The division methods and device of keyword | |
| CN109739958A (en) | A kind of specification handbook answering method and system |
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 |