CN113590797A - Intelligent operation and maintenance customer service system and implementation method - Google Patents
Intelligent operation and maintenance customer service system and implementation method Download PDFInfo
- Publication number
- CN113590797A CN113590797A CN202110898116.1A CN202110898116A CN113590797A CN 113590797 A CN113590797 A CN 113590797A CN 202110898116 A CN202110898116 A CN 202110898116A CN 113590797 A CN113590797 A CN 113590797A
- Authority
- CN
- China
- Prior art keywords
- intelligent
- customer service
- knowledge
- word
- module
- 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
- 238000012423 maintenance Methods 0.000 title claims abstract description 48
- 238000000034 method Methods 0.000 title claims abstract description 42
- 230000006854 communication Effects 0.000 claims abstract description 10
- 238000005516 engineering process Methods 0.000 claims abstract description 10
- 238000013473 artificial intelligence Methods 0.000 claims abstract description 5
- 238000004891 communication Methods 0.000 claims abstract description 4
- 230000011218 segmentation Effects 0.000 claims description 17
- 238000010276 construction Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000005494 condensation Effects 0.000 claims description 9
- 238000009833 condensation Methods 0.000 claims description 9
- 238000012216 screening Methods 0.000 claims description 6
- 238000013135 deep learning Methods 0.000 claims description 3
- 230000009191 jumping Effects 0.000 claims description 3
- 230000006698 induction Effects 0.000 claims description 2
- 230000008520 organization Effects 0.000 claims description 2
- 230000008569 process Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 101100261000 Caenorhabditis elegans top-3 gene Proteins 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000009411 base construction Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 239000007943 implant Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000008531 maintenance mechanism Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000004044 response Effects 0.000 description 1
Images
Classifications
-
- 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
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- 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
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- 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
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/216—Parsing using statistical methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/247—Thesauruses; Synonyms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- Mathematical Physics (AREA)
- Human Resources & Organizations (AREA)
- Databases & Information Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Operations Research (AREA)
- Molecular Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Computing Systems (AREA)
- Quality & Reliability (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Animal Behavior & Ethology (AREA)
- Biomedical Technology (AREA)
- Human Computer Interaction (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Machine Translation (AREA)
Abstract
The invention discloses an intelligent operation and maintenance customer service system, which comprises: one-key help module: the method comprises the steps that hot keys are integrated in a computer operating system and used for starting a client by one key; the intelligent customer service robot module is connected with the intelligent customer service robot module; intelligent customer service robot module: providing an instant communication environment for a user, extracting keyword information in a chat record by adopting an artificial intelligence technology, performing semantic recognition and intention recognition, and intelligently searching keywords in a knowledge base by adopting a word vector technology; is connected with the intelligent knowledge base module; the intelligent knowledge base module: providing a credit and creative industry intelligent dictionary library and a knowledge map for the intelligent customer service robot module to identify the intention of a user and search a required answer; the problems that the credit creation operation and maintenance service is long in time period, high in single service cost, low in efficiency and the like are effectively solved.
Description
Technical Field
The invention belongs to the technical field of intelligent operation and maintenance, and particularly relates to an intelligent operation and maintenance customer service system and an implementation method.
Background
With the improvement of the credit creation technology, a credit creation system gradually forms a free open ecology from an IT bottom layer framework and a standard, and ensures that information is independently controllable and safe; however, in the prior art, there is no mature operation and maintenance mechanism, especially a trusted operation and maintenance service, which mainly solves various technical problems encountered by a user in the process of using trusted equipment, including: the problems that the use habit of the user to the information creating device can not be developed step by step only by rapidly solving the problems that the use habit cannot be used, cannot be used or is not good, and the like, so that the use frequency is increased, and the information creating user can feel the service experience which is simple and efficient in time.
The current credit creation, operation and maintenance service forms mainly comprise active services and response services, the corresponding engineer service forms are generally door-to-door or telephone calls, and the requirements of credit creation users can be met in a general service scene, but if some difficult problems or unclear problems in the telephone calls are faced, the telephone service cannot achieve a good effect, and the door-to-door service forms the problems of long time period, high single service cost, low efficiency and the like, so that the more friendly and convenient service form is provided, which is particularly important.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the intelligent operation and maintenance customer service system and the implementation method are provided to solve the technical problems of long service cycle, high single cost, low efficiency and the like of the information, creation, operation and maintenance service in the prior art.
The technical scheme of the invention is as follows:
an intelligent operation and maintenance customer service system, comprising:
one-key help module: the method comprises the steps that hot keys are integrated in a computer operating system and used for starting a client by one key; the intelligent customer service robot module is connected with the intelligent customer service robot module;
intelligent customer service robot module: providing an instant communication environment for users, extracting keyword information in the chat records by adopting an artificial intelligence technology, carrying out intention identification, and intelligently matching keywords in a knowledge base by adopting a word vector technology. The intelligent robot module is connected with the intelligent knowledge base module;
the intelligent knowledge base module: and providing a credit and creative industry intelligent dictionary library and a knowledge map for the intelligent customer service robot module so as to identify the intention of the user and search the required answers.
The method is characterized in that an improved TF-IDF model is adopted to carry out word segmentation construction on technical documents, industry specifications and historical operation and maintenance documents of various manufacturers; and meanwhile, performing word segmentation according to a solution of a manual engineer, and updating the database of the intelligent dictionary database.
The knowledge graph is a letter creation knowledge graph constructed based on letter creation operation and maintenance documents.
It includes:
step 1, constructing an original dictionary through an improved TF-IDF model after word segmentation according to technical documents, industry specifications, standards and historical operation and maintenance documents provided by equipment related manufacturers in a letter-creation machine room and word segmentation according to a word segmentation algorithm;
step 2, dividing knowledge models related to the information creation equipment into three types: question and answer knowledge, article knowledge or documents; an initial credit creation knowledge map is constructed by combining the credit creation dictionary and the operation and maintenance document;
step 3, starting a client of the intelligent customer service robot module through a one-key help module;
step 4, according to the questions proposed by the user, performing intention recognition according to the original dictionary and the knowledge graph, and searching out articles, question-answering knowledge or documents corresponding to the keywords;
step 5, if the user cannot obtain the required answer, directly jumping to an engineer interface through an intelligent customer service robot module, solving the problem through manual intervention, and recording the communication process of the engineer;
and 6, extracting key words by adopting an improved TF-IDF model according to the records of the user and the manual communication process, taking the key words as a latest word bank, and updating the initial dictionary and the knowledge base.
The method for extracting the key words by adopting the TF-IDF model comprises the following steps:
step 6.1, dividing the text into single isolated words, removing stop words, and calculating the length of each isolated word as a candidate word;
and 6.2, calculating the TF value, the DF value, the degree of condensation and the degree of freedom of each word according to the obtained candidate word list, and screening a word bank meeting the requirements by calculating the TF-IDF value.
The calculation method of the TF-IDF value is as follows:
wherein t represents a candidate word, d represents a document, TF (t, d) represents the frequency of occurrence of the candidate word t in d, DF (t) represents the number of documents containing the candidate word t, D (t) represents the degree of cohesion of the candidate word t, and F (t) represents the degree of freedom of the candidate word t.
The calculation method of the condensation degree is as follows:
the degree of freedom is that the information entropy of the left and right adjacent characters of each candidate word is minimum, and the calculation mode is as follows:
F(t)＝min{p(ti-1)(-lg(p(ti-1))),p(ti+1)(-lg(p(ti+1)))}
where p (t) represents the probability that the term t appears in the current document as a whole, n represents the way in which the term t is combined in the current document, i-1 represents the left-adjacent word of the term t, and i +1 represents the right-adjacent word of the term t.
The construction method of the knowledge graph comprises the following steps: extracting entities, attributes and relations from data sources of question and answer knowledge, article knowledge, documents and an initial credit creation operation and maintenance dictionary by adopting an open-source Google knowledge map construction method, and adding the entities, the attributes and the relations into a data layer of a knowledge map; then induction organization is carried out, the concept is abstracted up gradually, and finally a mode layer is formed, so that the innovation knowledge graph is constructed.
The one-key help module starting method comprises the following steps: the method comprises the steps that a one-key help module is installed in terminal equipment through a sending channel, a hot key and a hot key calling program are written in through a system bottom layer interface, the intelligent customer service robot module program identifies different networks where the intelligent customer service robot module is located currently, a system default browser is opened through the system bottom layer interface, a robot page is accessed, a user clicks a set key on the terminal equipment, the default browser is called, and the intelligent customer service robot module automatically enters a conversation window.
The intention identification method comprises the steps of extracting keywords according to problems of a user by adopting an open-source deep learning algorithm, and searching a knowledge graph related to the keywords to serve as an intention identification result to be fed back to the user.
The invention has the beneficial effects that:
the invention adopts the technical scheme that firstly, in the intelligent operation and maintenance customer service system, a user can remotely control the system, and a help hot key is arranged on a keyboard to trigger service, so that the user can be ensured to call the service in the most direct and simple way; in the dictionary establishing stage, the condensation degree and the freedom degree parameters are fused, the TF-IDF model is improved, the problem that the TF-IDF model depends on an external text data set is effectively solved, and the initial dictionary establishing method can be used for establishing the initial dictionary for the document in the industry without depending on the external text data set (namely without any knowledge base). Meanwhile, the initial dictionary and the operation and maintenance related documents are combined to construct the knowledge map, so that the problems of synonyms, word ambiguity and the like in user semantic recognition are effectively solved. The intelligent operation and maintenance customer service system adopts the technologies of artificial intelligence, intention identification and the like, realizes self-service of users, responds to various service habits of the users, also helps to prolong a service time window, and effectively solves the problems of long time period, high single service cost, low efficiency and the like of the operation and maintenance service of the information and the innovation.
Drawings
FIG. 1 is a schematic diagram illustrating the principle of the intelligent operation and maintenance customer service system;
FIG. 2 is a schematic diagram of the logic principle of the intelligent operation and maintenance customer service system of the present invention;
FIG. 3 is a schematic flow chart of the intelligent operation and maintenance customer service system according to the present invention;
fig. 4 is a flow chart of the intelligent knowledge base construction of the intelligent operation and maintenance customer service system of the invention.
Detailed Description
An intelligent operation and maintenance customer service help system comprises a one-key help module, wherein a hot key is integrated in a computer operating system and used for starting a client by one key to quickly enter an intelligent customer service robot. The one-key help unit is connected with the intelligent customer service robot.
The intelligent customer service robot module provides an instant communication environment for users, adopts an artificial intelligence technology to extract keyword information in the chat records, performs intention identification, and adopts a word vector technology to intelligently match keywords in the knowledge base. And the intelligent robot module is connected with the intelligent knowledge base module.
The intelligent knowledge base module provides a credit and creative industry intelligent dictionary base and a knowledge map for the intelligent customer service robot, so that the intention of a user can be conveniently identified, and required answers can be quickly searched.
The intelligent dictionary database adopts an improved TF-IDF model to perform word segmentation on technical documents, industry specifications and historical operation and maintenance documents of various manufacturers, and constructs a credit creation operation and maintenance dictionary database. Meanwhile, word segmentation can be carried out according to the solution of a manual engineer, and a dictionary database is updated.
And the knowledge map is a letter creation knowledge map constructed based on letter creation operation and maintenance documents.
The implementation method of the intelligent operation and maintenance customer service help system specifically comprises the following steps:
step 1, according to technical documents provided by manufacturers related to equipment in a credit creation machine room and relevant industry specifications and standards, historical operation and maintenance documents, after word segmentation is carried out according to a word segmentation algorithm, original dictionary construction is carried out through an improved TF-IDF model;
step 2, dividing the model into three types according to the related knowledge related to the information: the method comprises the steps of establishing an initial credit creation knowledge map by combining question and answer knowledge, article knowledge and documents (attachments) with a credit creation dictionary and related operation and maintenance documents;
step 3, rapidly starting the intelligent robot module client through a one-key help module;
step 4, according to the questions proposed by the user, performing intention recognition according to the original dictionary and the knowledge graph, and searching out articles, question-answering knowledge and documents corresponding to the keywords;
step 5, if the user cannot obtain the required answer, directly jumping to an engineer interface through the intelligent robot module, solving the problem through manual intervention, and recording the communication process of the engineer;
and 6, extracting key words by adopting an improved TF-IDF model according to the records of the user and the manual communication process, taking the key words as a latest word bank, and updating the initial dictionary and the knowledge base.
The improved TF-IDF model is as follows: firstly, dividing a plurality of texts into single isolated words, removing stop words, and calculating the length of each isolated word as a candidate word; and calculating the TF value (namely the frequency of the word in the current text) and the DF value (the frequency of the word appearing in all documents), the Degree of condensation (Degree of condensation) and the Degree of freedom (freedom) of each word according to the obtained candidate word list, and screening a word bank meeting the requirements by calculating the TF-IDF value.
The calculation method of the TF-IDF value is as follows:
wherein t represents a candidate word, d represents a document, TF (t, d) represents the frequency of occurrence of the candidate word t in d, DF (t) represents the number of documents containing the candidate word t, D (t) represents the degree of cohesion of the candidate word t, and F (t) represents the information entropy of the candidate word t.
The degree of cohesion is defined as what part of the candidate words are combined, and the calculation method is as follows:
the degree of freedom is that the information entropy of the left and right adjacent characters of each candidate word is minimum, and the calculation mode is as follows:
F(t)＝min{p(ti-1)(-lg(p(ti-1))),p(ti+1)(-lg(p(ti+1)))}
where p (t) represents the probability that the term t appears in the current document as a whole, n represents the way in which the term t is combined in the current document, i-1 represents the left-adjacent word of the term t, and i +1 represents the right-adjacent word of the term t.
The knowledge graph construction method adopts an open-source Google knowledge graph construction method, extracts entities, attributes and relations from data sources of question and answer knowledge, article knowledge, documents (attachments) and an initial credit creation operation and maintenance dictionary, and adds the entities, the attributes and the relations into a data layer of the knowledge graph; then, the knowledge elements are generalized and organized, and are gradually abstracted upwards to concepts, and finally, a mode layer is formed, so that the innovation knowledge map is constructed.
The one-key help system module starting method comprises the following steps: the method comprises the steps that a one-key help system is installed in terminal equipment through a sending channel, hot keys are written in through a system bottom layer interface (files, commands and the like) and a hot key calling program (an intelligent robot program) is written in through the system bottom layer interface (files, commands and the like), the robot help program uses the system bottom layer interface (commands) to open a system default browser and access a robot page through recognizing different networks where the robot help program is located currently, a user clicks set keys on the terminal equipment, the default browser is called, and the intelligent robot conversation window is automatically entered.
The intention identification method adopts an open-source deep learning algorithm (such as LSTM, Bi-RNN, Bi-LSTM-CRF), extracts keywords according to the problems of the user, searches a knowledge graph related to the keywords, and selects a sentence pattern of Top3 as an intention identification result to feed back to the user.
The invention provides an intelligent operation and maintenance customer service help system and an implementation method thereof, which are starting points for comprehensive popularization of the credit industry in 2020, and aims to solve the problem that the prior art cannot realize intelligent operation and maintenance in the credit operation and maintenance process. The present invention will be described in detail below with reference to the accompanying drawings and examples.
The implementation process of the invention is as follows:
as shown in fig. 1 and fig. 3, an intelligent operation and maintenance customer service help system mainly includes three modules: the intelligent customer service system comprises a one-key help module, an intelligent customer service robot module and an intelligent knowledge base module, wherein the one-key help module is connected with the intelligent customer service robot module, and the intelligent customer service robot module is connected with the intelligent knowledge base module. The one-key help module firstly implants a hot key and a hot key calling program in a domestic operating system deployed by the information-creation equipment, the intelligent customer service robot module identifies different networks through the implanted program, a system bottom interface (command) is used for opening a system default browser, and a page of the intelligent operation and maintenance customer service robot is accessed. And the intelligent robot module is quickly matched to the solution in the knowledge base through the user intention analysis function. When the required solution cannot be searched in the knowledge base, the intelligent robot directly jumps to an online engineer until the problem is solved. The intelligent knowledge base firstly carries out word segmentation on technical documents, historical operation and maintenance information documents, industry standard specifications and other materials provided by various manufacturers to construct an initial dictionary, then classifies related documents, constructs an initial knowledge map by combining the initial dictionary database, subsequently carries out word segmentation according to a chat dialog box between a user and an engineer, and updates the dictionary and the knowledge base.
As shown in fig. 2, the invention is composed of hardware and software, the hardware is mainly a trusted cloud server, which provides the required computing, storage, network and database resources for the intelligent operation and maintenance system, and can be based on the user
The software mainly comprises a one-key help module, an intelligent robot module and an intelligent knowledge base module.
Wherein the cloud server configuration: hua is TaiShen 2280v2, CPU is more than or equal to 2 × 48Core @2.6GHz, memory is more than or equal to 24 × 32GB, SATA SSD is more than or equal to 2 × 480GB, SSD is more than or equal to 1 × 960GB, SATA is more than or equal to 3 × 4T, 8 × 25 GE;
operating the system: the Galaxy kylin operating system V10;
and (3) database configuration: wuhanda dream database.
As shown in fig. 4, the detailed intelligent knowledge base building process is that a user needs to manually sort out technical documents, industry specifications, historical operation and maintenance documents and other text data of various manufacturers, firstly, a Hidden Markov Model (HMM) word segmentation algorithm is adopted to segment a multi-text into single isolated words, stop words such as "and" are removed, and the length of each isolated word is calculated to serve as a candidate word; and calculating the TF value (namely the frequency of the word in the current text) and the DF value (the frequency of the word appearing in all documents), the Degree of condensation (Degree of condensation) and the Degree of Freedom (Freedom) of each word according to the obtained candidate word list, and screening a word bank meeting the requirements by calculating the TF-IDF value.
The first step is to calculate the values of the candidate words in TF and DF: TF is the number of occurrences of the candidate word in the document and DF is the number of occurrences of the candidate word in all documents.
The second step is to calculate the degree of cohesion D (t) and the degree of freedom F (t) of the candidate words:
F(t)＝min{p(ti-1)(-lg(p(ti-1))),p(ti+1)(-lg(p(ti+1)))}。
the third step is to calculate
And fourthly, setting stop words in the document by setting parameters of the stop words, if stop _ words is [ "yes" ], sorting TF-IDF-DOCF values of each candidate word, selecting 50 percent (specific parameters can be adjusted according to a screening result) of the top rank as the candidate words, and storing the candidate words into a Dameng database.
Fifthly, repeating TF-IDF-DOCF model retrieval according to the solution of a human engineer, selecting 90 percent (specific parameters can be adjusted according to the screening result) of the top rank as a new candidate word, and updating a credit creation dictionary database and a knowledge graph.
Claims (10)
1. An intelligent operation and maintenance customer service system, comprising:
one-key help module: the method comprises the steps that hot keys are integrated in a computer operating system and used for starting a client by one key; the intelligent customer service robot module is connected with the intelligent customer service robot module;
intelligent customer service robot module: providing an instant communication environment for users, extracting keyword information in the chat records by adopting an artificial intelligence technology, carrying out intention identification, and intelligently matching keywords in a knowledge base by adopting a word vector technology. The intelligent robot module is connected with the intelligent knowledge base module;
the intelligent knowledge base module: and providing a credit and creative industry intelligent dictionary library and a knowledge map for the intelligent customer service robot module so as to identify the intention of the user and search the required answers.
2. The intelligent operation and maintenance customer service system of claim 1, wherein: the construction method of the intelligent dictionary database comprises the following steps: the method is characterized in that an improved TF-IDF model is adopted to carry out word segmentation construction on technical documents, industry specifications and historical operation and maintenance documents of various manufacturers; and meanwhile, performing word segmentation according to a solution of a manual engineer, and updating the database of the intelligent dictionary database.
3. The intelligent operation and maintenance customer service system of claim 1, wherein: the knowledge graph is a letter creation knowledge graph constructed based on letter creation operation and maintenance documents.
4. The method of claim 1, wherein the method comprises: it includes:
step 1, constructing an original dictionary through an improved TF-IDF model after word segmentation according to technical documents, industry specifications, standards and historical operation and maintenance documents provided by equipment related manufacturers in a letter-creation machine room and word segmentation according to a word segmentation algorithm;
step 2, dividing knowledge models related to the information creation equipment into three types: question and answer knowledge, article knowledge or documents;
an initial credit creation knowledge map is constructed by combining the credit creation dictionary and the operation and maintenance document;
step 3, starting a client of the intelligent customer service robot module through a one-key help module;
step 4, according to the questions proposed by the user, performing intention recognition according to the original dictionary and the knowledge graph, and searching out articles, question-answering knowledge or documents corresponding to the keywords;
step 5, if the user cannot obtain the required answer, directly jumping to an engineer interface through an intelligent customer service robot module, solving the problem through manual intervention, and recording the communication process of the engineer;
and 6, extracting key words by adopting an improved TF-IDF model according to the records of the user and the manual communication process, taking the key words as a latest word bank, and updating the initial dictionary and the knowledge base.
5. The method of claim 4, wherein the method comprises: the method for extracting the key words by adopting the TF-IDF model comprises the following steps:
step 6.1, dividing the text into single isolated words, removing stop words, and calculating the length of each isolated word as a candidate word;
and 6.2, calculating the TF value, the DF value, the degree of condensation and the degree of freedom of each word according to the obtained candidate word list, and screening a word bank meeting the requirements by calculating the TF-IDF value.
6. The method of claim 5, wherein the method comprises:
the calculation method of the TF-IDF value is as follows:
wherein t represents a candidate word, d represents a document, TF (t, d) represents the frequency of occurrence of the candidate word t in d, DF (t) represents the number of documents containing the candidate word t, D (t) represents the degree of cohesion of the candidate word t, and F (t) represents the degree of freedom of the candidate word t.
7. The method of claim 5, wherein the method comprises:
the calculation method of the condensation degree is as follows:
the degree of freedom is that the information entropy of the left and right adjacent characters of each candidate word is minimum, and the calculation mode is as follows:
F(t)＝min{p(ti-1)(-lg(p(ti-1))),p(ti+1)(-lg(p(ti+1)))}
where p (t) represents the probability that the term t appears in the current document as a whole, n represents the way in which the term t is combined in the current document, i-1 represents the left-adjacent word of the term t, and i +1 represents the right-adjacent word of the term t.
8. The method of claim 5, wherein the method comprises: the construction method of the knowledge graph comprises the following steps: extracting entities, attributes and relations from data sources of question and answer knowledge, article knowledge, documents and an initial credit creation operation and maintenance dictionary by adopting an open-source Google knowledge map construction method, and adding the entities, the attributes and the relations into a data layer of a knowledge map; then induction organization is carried out, the concept is abstracted up gradually, and finally a mode layer is formed, so that the innovation knowledge graph is constructed.
9. The method of claim 5, wherein the method comprises: the one-key help module starting method comprises the following steps: the method comprises the steps that a one-key help module is installed in terminal equipment through a sending channel, a hot key and a hot key calling program are written in through a system bottom layer interface, the intelligent customer service robot module program identifies different networks where the intelligent customer service robot module is located currently, a system default browser is opened through the system bottom layer interface, a robot page is accessed, a user clicks a set key on the terminal equipment, the default browser is called, and the intelligent customer service robot module automatically enters a conversation window.
10. The method of claim 5, wherein the method comprises: the intention identification method comprises the steps of extracting keywords according to problems of a user by adopting an open-source deep learning algorithm, and searching a knowledge graph related to the keywords to serve as an intention identification result to be fed back to the user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110898116.1A CN113590797A (en) | 2021-08-05 | 2021-08-05 | Intelligent operation and maintenance customer service system and implementation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110898116.1A CN113590797A (en) | 2021-08-05 | 2021-08-05 | Intelligent operation and maintenance customer service system and implementation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113590797A true CN113590797A (en) | 2021-11-02 |
Family
ID=78255643
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110898116.1A Pending CN113590797A (en) | 2021-08-05 | 2021-08-05 | Intelligent operation and maintenance customer service system and implementation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113590797A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI807569B (en) * | 2022-01-03 | 2023-07-01 | 中華電信股份有限公司 | Auto-reply system and auto-reply method based on instant message service |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110209787A (en) * | 2019-05-29 | 2019-09-06 | 袁琦 | A kind of intelligent answer method and system based on pet knowledge mapping |
| CN110232573A (en) * | 2018-03-06 | 2019-09-13 | 广州供电局有限公司 | Based on interactive intelligent response system |
| CN110413761A (en) * | 2019-08-06 | 2019-11-05 | 浩鲸云计算科技股份有限公司 | A kind of method that the territoriality in knowledge based library is individually talked with |
| CN110457708A (en) * | 2019-08-16 | 2019-11-15 | 腾讯科技（深圳）有限公司 | Vocabulary mining method, apparatus, server and storage medium based on artificial intelligence |
| CN111291156A (en) * | 2020-01-21 | 2020-06-16 | 同方知网(北京)技术有限公司 | Question-answer intention identification method based on knowledge graph |
| CN111309877A (en) * | 2018-12-12 | 2020-06-19 | 北京文因互联科技有限公司 | Intelligent question-answering method and system based on knowledge graph |
| WO2020237856A1 (en) * | 2019-05-29 | 2020-12-03 | 平安科技（深圳）有限公司 | Smart question and answer method and apparatus based on knowledge graph, and computer storage medium |
| CN112597760A (en) * | 2020-12-04 | 2021-04-02 | 光大科技有限公司 | Method and device for extracting domain words in document |
-
2021
- 2021-08-05 CN CN202110898116.1A patent/CN113590797A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110232573A (en) * | 2018-03-06 | 2019-09-13 | 广州供电局有限公司 | Based on interactive intelligent response system |
| CN111309877A (en) * | 2018-12-12 | 2020-06-19 | 北京文因互联科技有限公司 | Intelligent question-answering method and system based on knowledge graph |
| CN110209787A (en) * | 2019-05-29 | 2019-09-06 | 袁琦 | A kind of intelligent answer method and system based on pet knowledge mapping |
| WO2020237856A1 (en) * | 2019-05-29 | 2020-12-03 | 平安科技（深圳）有限公司 | Smart question and answer method and apparatus based on knowledge graph, and computer storage medium |
| CN110413761A (en) * | 2019-08-06 | 2019-11-05 | 浩鲸云计算科技股份有限公司 | A kind of method that the territoriality in knowledge based library is individually talked with |
| CN110457708A (en) * | 2019-08-16 | 2019-11-15 | 腾讯科技（深圳）有限公司 | Vocabulary mining method, apparatus, server and storage medium based on artificial intelligence |
| CN111291156A (en) * | 2020-01-21 | 2020-06-16 | 同方知网(北京)技术有限公司 | Question-answer intention identification method based on knowledge graph |
| CN112597760A (en) * | 2020-12-04 | 2021-04-02 | 光大科技有限公司 | Method and device for extracting domain words in document |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI807569B (en) * | 2022-01-03 | 2023-07-01 | 中華電信股份有限公司 | Auto-reply system and auto-reply method based on instant message service |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11556713B2 (en) | System and method for performing a meaning search using a natural language understanding (NLU) framework | |
| TWI732271B (en) | Human-machine dialog method, device, electronic apparatus and computer readable medium | |
| CN110852086B (en) | Artificial intelligence based ancient poetry generating method, device, equipment and storage medium | |
| CN110032632A (en) | Intelligent customer service answering method, device and storage medium based on text similarity | |
| WO2021121198A1 (en) | Semantic similarity-based entity relation extraction method and apparatus, device and medium | |
| WO2020082560A1 (en) | Method, apparatus and device for extracting text keyword, as well as computer readable storage medium | |
| CN111241237B (en) | Intelligent question-answer data processing method and device based on operation and maintenance service | |
| CN111274365B (en) | Intelligent inquiry method and device based on semantic understanding, storage medium and server | |
| CN108304372A (en) | Entity extraction method and apparatus, computer equipment and storage medium | |
| CN111931500B (en) | Search information processing method and device | |
| CN111291195B (en) | Data processing method, device, terminal and readable storage medium | |
| CN110287325A (en) | A kind of power grid customer service recommended method and device based on intelligent sound analysis | |
| CN111651572A (en) | Multi-domain task type dialogue system, method and terminal | |
| CN111061837A (en) | Topic identification method, device, equipment and medium | |
| CN108304424A (en) | Text key word extracting method and text key word extraction element | |
| KR101545050B1 (en) | Method for automatically classifying answer type and apparatus, question-answering system for using the same | |
| CN113779987A (en) | Event co-reference disambiguation method and system based on self-attention enhanced semantics | |
| CN113590797A (en) | Intelligent operation and maintenance customer service system and implementation method | |
| CN109472032A (en) | A kind of determination method, apparatus, server and the storage medium of entity relationship diagram | |
| WO2021139076A1 (en) | Intelligent text dialogue generation method and apparatus, and computer-readable storage medium | |
| CN112417875A (en) | Configuration information updating method and device, computer equipment and medium | |
| CN110874408B (en) | Model training method, text recognition device and computing equipment | |
| CN113761081A (en) | Method and system for carrying out multi-dimensional combined retrieval on enterprise information | |
| CN111460106A (en) | Information interaction method, device and equipment | |
| CN111767377B (en) | Efficient spoken language understanding and identifying method oriented to low-resource environment |
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: 20211102 |
|
| RJ01 | Rejection of invention patent application after publication |