CN109189906A - Intelligent customer service is to the complete semantic recognition methods of more question sentences under coherent context - Google Patents
Intelligent customer service is to the complete semantic recognition methods of more question sentences under coherent context Download PDFInfo
- Publication number
- CN109189906A CN109189906A CN201810939124.4A CN201810939124A CN109189906A CN 109189906 A CN109189906 A CN 109189906A CN 201810939124 A CN201810939124 A CN 201810939124A CN 109189906 A CN109189906 A CN 109189906A
- Authority
- CN
- China
- Prior art keywords
- words
- question sentence
- user
- matched
- word
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention relates to intelligent customer services to the complete semantic recognition methods of more question sentences under coherent context.It comprises the following processes: step 1: rejecting stop words, formed " user's question sentence set of words "；Step 2: to first question sentence, finding out the matching highest standard question sentence of degree；The user's word matched is saved as into " set of words consumed "；Step 3: to next question sentence, being matched；If can match, word will be matched and be added to " set of words consumed ", while return step 5；If not completing matching, 4 are gone to step；Step 4: being supplemented from " set of words consumed "；Step 5: judging whether user's question sentence set of words also has residue, if there is residue, return to 3, otherwise go to step 6；Step 6: all complete standard question sentences of output.The present invention can be solved the problems, such as effectively that semantic information can not be shared in punctuate, so as to shorten turnaround time, improve and reply accuracy rate and system intelligence.
Description
Technical field:
The present invention relates to intelligent customer service technologies, further to intelligent customer service to the complete semanteme of more question sentences under coherent context
Recognition methods.
Background technique:
95598 business for accepting the whole network are concentrated for support Guo Wang client service center, ensure that level of customer service is persistently promoted, state
Net client service center builds in September, 2013 and has put into operation Customer Service Center of State Grid Corporation of China knowledge base system (hereinafter referred to as
" center knowledge bases ").Center knowledge bases are guiding with knowledge services, establish Knowledge Management System and framework, construct unification
Customer service knowledge hierarchy and knowledge application, the intelligence provided fast, accurately and comprehensively towards Guo Wang client service center customer service officer are known
Know service, effectively shorten response time of customer service officer, improves disposable answer rate and business replies accuracy, promote visitor
Family meets Guo Wang Customer Service Center and externally provides unified Customer Service Information branch to the total satisfaction of hot-line service
The requirement of support.
More than the 4000 customer service officer trainings of existing knowledge library system main support Guo Wang client service center, study, scene operation.
Current knowledge collection process initiates knowledge altogether and collects process 81707, creation of knowledge catalogue 3128, compiles Knowledge Element
It is 108233, average day retrieval nearly 30,000 times, average daily to read 4.8 ten thousand, it is played an important role for 95598 customer services work.
In the system of the customer service of electric system, the visiting knowledge base system of client is solved the problems, such as, gradually from passively
Business tine retrieval redirect to automation, the Collaborative operational support system for more actively, being more bonded business service process.
Human-computer interaction (Human-Computer Interaction, HCI) is to interact pass between research system and user
The science of system.System can be various machines, be also possible to the system and software of computerization.For example, by man-machine
The various artificial intelligence systems such as intelligent customer service system, speech control system may be implemented in interaction.
Intelligent Answer System is a kind of typical case of human-computer interaction.Traditional intelligent Answer System is to propose user
Problem directly carries out similarity calculation with a large amount of problems stored in knowledge base, obtains the answer to match with the problem.Due to
The customer service business of electric system, itself in face of person service object population-wide is wide, device model is various, failure cause is more,
Failure solves a series of problems, such as more complex, to need to build the business process map of magnanimity, and numerous involved in electric service
Business datum platform.Therefore, calculation amount is very big, causes computational efficiency low.In addition, the above method can only ask single intention
Topic or the more intention problems that can effectively make pauses in reading unpunctuated ancient writings are replied, and the accuracy rate that answer is replied is low, therefore, leads to user experience
Difference.
Summary of the invention:
Main services direct object of the invention is using the Electricity customers of supply intelligent customer service robot, indirect one pair
As the online customer service personnel for national grid client service center, when the complicated business scene problem encountered for them, to problem
Solution, service steps process are in use, the intelligent auxiliary carried out.
Relational language is explained:
Standard is asked: the text for indicating some knowledge point, and main target is that expression is clear, convenient for safeguarding.As " CRBT
Rate " are exactly that clearly standard asks description for expression.
Extension is asked: semantic formula and natural sentence set for indicating some knowledge point semanteme.
Semantic formula: semantic formula is mainly made of word, part of speech and their "or" relationship, and core depends on
" part of speech ", part of speech simply understand to be one group of word for having general character, these words can be similar or dissimilar semantically, this
A little words can also be noted as important or inessential.Semantic formula and user's question sentence relationship and traditional template matching have very
Big difference, in conventional template matching, template and user's question sentence only match with not matched relationship, and semantic formula with
Relationship is indicated by the value (similarity) of quantization between user's question sentence, while value and similar question sentence and the user of this quantization
Similarity between question sentence can mutually compare.Since semantic formula will participate in together similarity meter with similar question sentence
It calculates, so the definition of template grammar is unsuitable complicated, but has enough ability expression semantic again.
Context relation question and answer: when lacking some key messages in when the user the problem of, robot can be by such problems
With user above in association with getting up to analyze, then provide most suitable answer.There is associations between some specific knowledge points
Relationship, robot has the ability of " memory user is above ", and can be combined together to be parsed and provided with new problem and answer
Case embodies the fluency and intelligence of robot question answering process.
Specific technical solution is as follows:
Intelligent customer service is text formatting to the complete semantic recognition methods of more question sentences under coherent context, more question sentences；Including such as
Lower process:
Step 1: rejecting prefix word, suffix word, stop words, participle in text, formed " user's question sentence set of words "；
Step 2: according to sequence from left to right, to first question sentence, finding out in knowledge base and matched with first sentence
Spend highest standard question sentence；The user's word matched is saved as into " set of words consumed "；
Step 3: to next question sentence, being matched；It, will matching if can be matched with the standard question sentence in knowledge base
Complete user's word is added to " set of words consumed " in step 2, while return step 5；If not completing matching, turn
Step 4；
Step 4: which word lacked in matching, is supplemented from " set of words consumed ", and update and " disappeared
The set of words that expense is fallen "；
Step 5: judging whether user's question sentence set of words also has residue, if there is residue, return to 3, otherwise go to step 6；
Step 6: all complete standard question sentences of output.
It before step 1 further include a preposition step when the problem of user is other formats in addition to text formatting
Rapid 0, step 0: format the problem of user is converted into text formatting.
The progress of the present invention compared with the existing technology is: the present invention is to existing Power System Intelligent customer service question answering system
Knowledge architecture method, process and content do not increase any extra work, pass through the similarity asked customer problem and extension
Calculate, it can be achieved that under Same Scene continuous question sentence understanding, and provide the corresponding answer of each problem；The present invention can be solved effectively
The problem of semantic information can not be shared in punctuate, improves so as to shorten turnaround time and replys accuracy rate and system intelligence.
Detailed description of the invention:
Fig. 1 is the recognition methods flow diagram that intelligent customer service is complete semantic to more question sentences under coherent context in embodiment.
Specific embodiment:
Embodiment:
Is as client asks: how intelligent electric meter replaced? does is it free? whom affiliated people? how to supplement with money?
Since multiple intentions in model sentence have shared semantic information, institute not can be carried out effectively disconnected in the conventional way
Sentence.
The knowledge point of knowledge base is as follows:
Is (1) how intelligent electric meter replaced?
Expression formula: [intelligent electric meter] [how] [replacement]
(1) whether intelligent electric meter replacement free?
Expression formula: [intelligent electric meter] [replacement] [whether] [it is free | charge]
Does is (3) whom the affiliated people of intelligent electric meter?
Expression formula: [intelligent electric meter] [affiliated people] [being] [who]
Does (4) how intelligent electric meter supplement with money?
Expression formula: [intelligent electric meter] [how] [supplementing with money]
Is when user asks in a question sentence: how intelligent electric meter replaced? does is it free? whom affiliated people? how to fill
Value?
According to process: above four expression formulas are all Candidate Set.
Step 1: rejecting prefix word, suffix word, stop words, participle in text, formed " user's question sentence set of words ", it may be assumed that [intelligence
Energy ammeter] [how] [replacement] [being] [free] [affiliated people] [being] [who] [how] [supplementing with money]；
Step 2: according to sequence from left to right, to first question sentence, finding out in knowledge base and matched with first sentence
Spend highest standard question sentence；The user's word matched is saved as into " set of words consumed "；First matched expression
Formula are as follows: [intelligent electric meter] [how] [replacement], [set of words consumed ] are as follows: intelligent electric meter, how, replacement；
Step 3: to next question sentence, being matched；It, will matching if can be matched with the standard question sentence in knowledge base
Complete user's word is added to " set of words consumed " in step 2, while return step 5；If not completing matching, turn
Step 4；
Step 4: which word lacked in matching, is supplemented from " set of words consumed ", and update and " disappeared
The set of words that expense is fallen "；
In the example, by the consumption principle of user's question sentence from left to right, extension, which is asked, to meet the requirements: [intelligent electric meter] [replacement]
[whether] [free | charge], but the expression formula lacks two necessary words: intelligent electric meter replacement, then from [having been consumed
Set of words] in find, and the expression formula consumed complete；
Step 5: judging whether user's question sentence set of words also has residue, if there is residue, return to 3, otherwise go to step 6；
Step 6: all complete standard question sentences of output.
Claims (2)
1. intelligent customer service is to the complete semantic recognition methods of more question sentences under coherent context, which is characterized in that more question sentences are text lattice
Formula；It comprises the following processes:
Step 1: rejecting prefix word, suffix word, stop words, participle in text, formed " user's question sentence set of words "；
Step 2: according to sequence from left to right, to first question sentence, found out in knowledge base with first sentence matching degree most
High standard question sentence；The user's word matched is saved as into " set of words consumed "；
Step 3: to next question sentence, being matched；If can be matched with the standard question sentence in knowledge base, by what is matched
User's word is added to " set of words consumed " in step 2, while return step 5；If not completing matching, go to step
4；
Step 4: which word lacked in matching, is supplemented from " set of words consumed ", matched after update " by
The set of words consumed "；
Step 5: judging whether user's question sentence set of words also has residue, if there is residue, return to 3, otherwise go to step 6；
Step 6: all complete standard question sentences of output.
2. intelligent customer service is to the complete semantic recognition methods of more question sentences under coherent context, which is characterized in that comprise the following processes:
Step 0: format the problem of user is converted into text formatting；
Step 1: rejecting prefix word, suffix word, stop words, participle in text, formed " user's question sentence set of words "；
Step 2: according to sequence from left to right, to first question sentence, found out in knowledge base with first sentence matching degree most
High standard question sentence；The user's word matched is saved as into " set of words consumed "；
Step 3: to next question sentence, being matched；If can be matched with the standard question sentence in knowledge base, by what is matched
User's word is added to " set of words consumed " in step 2, while return step 5；If not completing matching, go to step
4；
Step 4: which word lacked in matching, is supplemented from " set of words consumed ", and update and " consumed
Set of words "；
Step 5: judging whether user's question sentence set of words also has residue, if there is residue, return to 3, otherwise go to step 6；
Step 6: all complete standard question sentences of output.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810939124.4A CN109189906A (en) | 2018-08-17 | 2018-08-17 | Intelligent customer service is to the complete semantic recognition methods of more question sentences under coherent context |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810939124.4A CN109189906A (en) | 2018-08-17 | 2018-08-17 | Intelligent customer service is to the complete semantic recognition methods of more question sentences under coherent context |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109189906A true CN109189906A (en) | 2019-01-11 |
Family
ID=64918646
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810939124.4A Pending CN109189906A (en) | 2018-08-17 | 2018-08-17 | Intelligent customer service is to the complete semantic recognition methods of more question sentences under coherent context |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109189906A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110008322A (en) * | 2019-03-25 | 2019-07-12 | 阿里巴巴集团控股有限公司 | Art recommended method and device under more wheel session operational scenarios |
| CN111831801A (en) * | 2020-05-27 | 2020-10-27 | 北京市农林科学院 | Man-machine conversation method and system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104573028A (en) * | 2015-01-14 | 2015-04-29 | 百度在线网络技术（北京）有限公司 | Intelligent question-answer implementing method and system |
| US20150220626A1 (en) * | 2014-01-31 | 2015-08-06 | Verint Systems Ltd. | Automated Removal of Private Information |
| CN106528759A (en) * | 2016-11-04 | 2017-03-22 | 上海智臻智能网络科技股份有限公司 | Information processing method and device for intelligent question-answering system |
| CN106649825A (en) * | 2016-12-29 | 2017-05-10 | 上海智臻智能网络科技股份有限公司 | Voice interaction system, establishment method and device thereof |
| CN107273350A (en) * | 2017-05-16 | 2017-10-20 | 广东电网有限责任公司江门供电局 | A kind of information processing method and its device for realizing intelligent answer |
| CN107967261A (en) * | 2017-11-17 | 2018-04-27 | 康成投资（中国）有限公司 | Interactive question semanteme understanding method in intelligent customer service |
-
2018
- 2018-08-17 CN CN201810939124.4A patent/CN109189906A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150220626A1 (en) * | 2014-01-31 | 2015-08-06 | Verint Systems Ltd. | Automated Removal of Private Information |
| CN104573028A (en) * | 2015-01-14 | 2015-04-29 | 百度在线网络技术（北京）有限公司 | Intelligent question-answer implementing method and system |
| CN106528759A (en) * | 2016-11-04 | 2017-03-22 | 上海智臻智能网络科技股份有限公司 | Information processing method and device for intelligent question-answering system |
| CN106649825A (en) * | 2016-12-29 | 2017-05-10 | 上海智臻智能网络科技股份有限公司 | Voice interaction system, establishment method and device thereof |
| CN107273350A (en) * | 2017-05-16 | 2017-10-20 | 广东电网有限责任公司江门供电局 | A kind of information processing method and its device for realizing intelligent answer |
| CN107967261A (en) * | 2017-11-17 | 2018-04-27 | 康成投资（中国）有限公司 | Interactive question semanteme understanding method in intelligent customer service |
Non-Patent Citations (2)
| Title |
|---|
| SIMON A 等: "Holographic entanglement and causal information in coherent states", 《JOURNAL OF HIGH ENERGY PHYSICS 2014》 * |
| 董自涛 等: "智能问答系统中问句相似度计算方法", 《武汉理工大学学报（信息与管理工程版）》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110008322A (en) * | 2019-03-25 | 2019-07-12 | 阿里巴巴集团控股有限公司 | Art recommended method and device under more wheel session operational scenarios |
| CN110008322B (en) * | 2019-03-25 | 2023-04-07 | 创新先进技术有限公司 | Method and device for recommending dialogues in multi-turn conversation scene |
| CN111831801A (en) * | 2020-05-27 | 2020-10-27 | 北京市农林科学院 | Man-machine conversation method and system |
| CN111831801B (en) * | 2020-05-27 | 2022-02-25 | 北京市农林科学院 | Man-machine conversation method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA2929018C (en) | Natural expression processing method, processing and response method, device and system | |
| CN108470282A (en) | Work order intelligent method for classifying is complained by Customer Service Center | |
| CN112507700A (en) | Event extraction method and device, electronic equipment and storage medium | |
| CN104008160A (en) | Method and system of indistinct logic chatting robot for realizing parallel topic control | |
| CN110781663A (en) | Training method and device of text analysis model and text analysis method and device | |
| CN108763212A (en) | A kind of address information extraction method and device | |
| CN108510292A (en) | Automatic flow householder method for fault scenes problem in electric power calling service | |
| CN109189906A (en) | Intelligent customer service is to the complete semantic recognition methods of more question sentences under coherent context | |
| CN106127526A (en) | Intelligent robot system and method for work thereof | |
| CN115309877A (en) | Dialog generation method, dialog model training method and device | |
| Bunt | A context-change semantics for dialogue acts | |
| CN113360001A (en) | Input text processing method and device, electronic equipment and storage medium | |
| CN113761927B (en) | Power grid fault handling real-time auxiliary decision-making method, system, equipment and storage medium | |
| CN112801721B (en) | Information processing method, information processing device, electronic equipment and storage medium | |
| CN114020886A (en) | Speech intention recognition method, device, equipment and storage medium | |
| CN117194624A (en) | Traffic data analysis complex task intelligent disassembling and completing method based on large language model | |
| Ou et al. | Automatic Multimedia-based Question-Answer Pairs Generation in Computer Assisted Healthy Education System | |
| CN114328959B (en) | Knowledge graph construction and use method, device, equipment and medium | |
| CN106407271B (en) | Intelligent customer service system and updating method of intelligent customer service knowledge base thereof | |
| CN114220425A (en) | Chat robot system and conversation method based on voice recognition and Rasa framework | |
| CN114239583A (en) | Method, device, equipment and medium for training entity chain finger model and entity chain finger | |
| CN109543182A (en) | A kind of electric power enterprise based on solr engine takes turns interactive semantic analysis method more | |
| CN112380332A (en) | Interactive knowledge feedback method, device and computer storage medium | |
| CN112685434A (en) | Operation and maintenance question-answering method based on knowledge graph | |
| Karoui et al. | Semi-automatic domain ontology construction from spoken corpus in Tunisian dialect: Railway request information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190111 |
|
| RJ01 | Rejection of invention patent application after publication |