CN109062896A - A kind of matching process and system based on artificial intelligence words art model - Google Patents
A kind of matching process and system based on artificial intelligence words art model Download PDFInfo
- Publication number
- CN109062896A CN109062896A CN201810825399.5A CN201810825399A CN109062896A CN 109062896 A CN109062896 A CN 109062896A CN 201810825399 A CN201810825399 A CN 201810825399A CN 109062896 A CN109062896 A CN 109062896A
- Authority
- CN
- China
- Prior art keywords
- information
- words art
- name entity
- art model
- artificial intelligence
- 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
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
- G06F40/295—Named entity recognition
-
- 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 invention discloses a kind of matching process and system based on artificial intelligence words art model, and method includes: S1, input information according to user, keyword extraction inputs the name entity in information;S2 obtains corresponding informance position of the name entity in words art model according to the name entity;S3 obtains knowledge information corresponding with the information of input according to the information position;S4 generates corresponding response content according to the knowledge information.The present invention conversation content different for user, the name entity in conversation content is extracted using keyword or crucial sentence pattern, it is matched in words art model and obtains knowledge information using deep learning neural network, generate corresponding response content, realize the reciprocity in human-computer dialogue, so as to interact machine with user more naturally, the interactivity and user experience of dialogue are promoted.
Description
Technical field
The present invention relates to field of artificial intelligence more particularly to a kind of matching process based on artificial intelligence words art model
And system.
Background technique
In recent years, the rise of artificial intelligence is so that the artificial intelligence system of interactive question and answer achieves important breakthrough, but shows
Some interactive request-answering systems are in the communication process with user, and still seem inadequate " intelligence ".
In existing interactive system, mostly uses the mode in knowledge based library to generate alternative answer greatly, then answer again
Case is replied after being ranked up, but this match pattern, be easy to cause matching chaotic, machine is given an irrelevant answer, so that human-computer interaction is not
Natural, discontinuous, user experience is not good enough.
Summary of the invention
The purpose of the present invention is to provide a kind of matching process and system based on artificial intelligence words art model, it is therefore intended that
Solve the problems, such as that human-computer dialogue lacks initiative and is unable to satisfy user demand in existing artificial intelligence system.
To achieve the above object, technical scheme is as follows:
A kind of matching process based on artificial intelligence words art model, includes the following steps,
S1 inputs information according to user, and keyword extraction inputs the name entity in information;
S2 obtains corresponding informance position of the name entity in words art model according to the name entity;
S3 obtains knowledge information corresponding with input information according to the information position;
S4 generates corresponding response content according to the knowledge information.
In above scheme, the step S1 is specifically included: natural language processing technique and semantic analysis technology are used, it is crucial
Word or crucial sentence pattern obtain in user's input information and name entity.
In above scheme, the step S2 is specifically included: being preset multiple matching position point locations in words art model, is utilized
The number that the name entity of extraction occurs, the corresponding position where being matched in words art model.
In above scheme, the step S3 is specifically included: according to shown corresponding informance position, utilizing deep learning nerve net
Network obtains knowledge information, including general knowledge information and professional knowledge letter after arranging by data mining to dialog information
Breath.
In above scheme, the step S4 is specifically included: according to the knowledge information, using deep learning neural network
Ability in feature extraction extracts the relationship that user inputs between the feature of information and name entity attribute, determines response content, output
With result.
A kind of matching system based on artificial intelligence words art model, comprising:
Entity extraction unit inputs the name entity in information for keyword extraction user;
Model Matching unit, for determining position of the name entity in words art model;
Knowledge extraction unit, for extracting name entity in the various knowledge informations of words art modal position information;
User is inputted the feature of problem and the knowledge of name entity in information by deep learning and believed by information answer unit
Manner of breathing association, determines final reply.
In above scheme, the words art model is a kind of database based on figure, and the content of storage is name entity, each
It names associated by attribute between entity.
Of the invention matching process and system based on artificial intelligence words art model, for the different conversation content of user,
The name entity in conversation content is extracted using keyword or crucial sentence pattern, be matched in words art model and utilizes deep learning mind
Knowledge information is obtained through network, generates corresponding response content, realizes the reciprocity in human-computer dialogue, so as to make machine more certainly
It so is interacted with user, promotes the interactivity and user experience of dialogue.
Detailed description of the invention
Fig. 1 is the flow chart of the matching process based on artificial intelligence words art model of one embodiment of the invention;
Fig. 2 is the structural schematic diagram of the matching system based on artificial intelligence words art model of one embodiment of the invention.
Specific embodiment
Technical solution of the present invention is described in further detail with reference to the accompanying drawings and examples.
Fig. 1 is the flow chart of the matching process based on artificial intelligence words art model of one embodiment of the invention.
As shown in Figure 1, should go out name entity by keyword extraction based on the matching process of artificial intelligence words art model,
The position that name entity is navigated in words art information system, obtains the corresponding informance of name entity, using deep learning
The ability in feature extraction of neural network extracts the relationship that user inputs between the feature of question sentence and name entity attribute, thus really
Fixed corresponding response content, specifically includes following steps:
S1 inputs information according to user, the name entity in keyword extraction information.
User is obtained using natural language processing technique and semantic analysis technology and inputs the keyword of information or using nature
Language processing techniques and semantic analysis technology obtain the crucial sentence pattern that user inputs information;
Such as: user input " which country capital Washington be? " according to input information extraction " Washington " as name
Entity.
Wherein, natural language processing technique uses Seq2Seq method, covers not covered in original application ask automatically
It answers, by big data and upgrades continuous error correction upgrading automatically and carried out based on Bi-LSTM technology by multilayer deep neural network
Mood and intention judgement;Semantic analysis technology studies the meaning of things with semantic differential scale, using several 7 grades
Semantic scale name entity that keyword or crucial sentence pattern are extracted evaluate, determine in each meaning commented in dimension and
Intensity.
S2 obtains corresponding informance position of the entity in words art model according to the name entity;
Words art model is a kind of database based on figure, and the content of storage is name entity, is led between each name entity
It is associated to cross attribute.Multiple matching position point locations are preset in words art model, are gone out using the entity for obtaining user's input information
Existing number, the corresponding position where being matched to entity in words art model.
Such as: according to " New York " as name entity, " Washington " is oriented from words art model and is talking about letter in art model
Cease position.
S3 obtains knowledge information corresponding with the information of input according to the information position.
According to shown dialog information position, using deep learning neural network, dialog information is carried out by data mining
Knowledge information, including general knowledge information and expert knowledge information are obtained after arrangement, wherein general knowledge information is common-sense
Knowledge information.
Each attribute of name entity, such as Washington-capital-U.S., Washington-position-are obtained according to information position
The knowledge informations such as North America.
S4 generates corresponding response content according to the knowledge information.
According to the knowledge information, the feature of question sentence and the relationship of each attributive character are extracted by deep learning method,
To obtaining corresponding response content, it is determined that user inputs information " which country capital Washington be? " in " capital " letter
Breath is more related to problem, and final response content is " U.S. ".
In order to realize above-described embodiment, the present invention also provides a kind of matching systems based on artificial intelligence words art model.
Fig. 2 is the structural schematic diagram of the matching system based on artificial intelligence words art model of one embodiment of the invention.
As shown in Fig. 2, should be based on the matching system of artificial intelligence words art model, including entity extraction unit, Model Matching
Unit, knowledge extraction unit and information answer unit, wherein
Entity extraction unit inputs the name entity in information for keyword extraction user;
Model Matching unit, for determining position of the name entity in words art model;
Knowledge extraction unit, for extracting name entity in each attribute knowledge information of words art modal position information;
Information answer unit is known the feature of problem in user's input information with name entity attributes by deep learning
It is associated to know information, determines final reply.
Such as: when user input information be " which country capital Washington be? "
By entity extraction unit, it is real as name that " Washington " in Information Problems is inputted using keyword extraction user
Body;
Go out to name the information position of entity " Washington " in words art model in Model Matching units match;
Name entity " Washington " is extracted in each attribute knowledge of words art model corresponding position according to knowledge extraction unit
Information, such as Washington-capital-U.S., Washington-position-North America etc.;
User is extracted finally by deep learning method and inputs the feature of question sentence and the relationship of each attributive character, to obtain
Corresponding response content is taken, final to determine that " capital " information is more related to problem, information answer unit determines final reply " beauty
State ".
Of the invention matching process and system based on artificial intelligence words art model, for the different conversation content of user,
The name entity in conversation content is extracted using keyword or crucial sentence pattern, be matched in words art model and utilizes deep learning mind
Knowledge information is obtained through network, generates corresponding response content, realizes the reciprocity in human-computer dialogue, so as to make machine more certainly
It so is interacted with user, promotes the interactivity and user experience of dialogue.
Above-described specific embodiment has carried out further the purpose of the present invention, technical scheme and beneficial effects
It is described in detail, it should be understood that the foregoing is merely a specific embodiment of the invention, the guarantor that is not intended to limit the present invention
Range is protected, all within the spirits and principles of the present invention, any modification, equivalent substitution, improvement and etc. done should all be contained in this hair
Within bright protection scope.
Claims (7)
1. a kind of matching process based on artificial intelligence words art model, which comprises the steps of:
S1 inputs information according to user, and keyword extraction inputs the name entity in information;
S2 obtains corresponding informance position of the name entity in words art model according to the name entity;
S3 obtains knowledge information corresponding with input information according to the information position;
S4 generates corresponding response content according to the knowledge information.
2. the matching process according to claim 1 based on artificial intelligence words art model, it is characterised in that: the step S1
Include:
Using natural language processing technique and semantic analysis technology, keyword or crucial sentence pattern are obtained and are named in user's input information
Entity.
3. the matching process according to claim 1 based on artificial intelligence words art model, it is characterised in that: the step S2
Include:
Multiple matching position point locations are preset in words art model, the number occurred using the name entity of extraction is matched
Corresponding position into place words art model;The words art model is a kind of database based on figure, and the content of storage is name
Entity, it is associated by attribute between each name entity.
4. the matching process according to claim 1 based on artificial intelligence words art model, it is characterised in that: the step S3
Include:
According to shown corresponding informance position, using deep learning neural network, dialog information is arranged by data mining
After obtain knowledge information, including general knowledge information and expert knowledge information.
5. the matching process according to claim 1 based on artificial intelligence words art model, it is characterised in that: the step S4
Include:
The spy that user inputs information is extracted using the ability in feature extraction of deep learning neural network according to the knowledge information
Relationship between sign and name entity attribute, determines response content, exports matching result.
6. a kind of matching system based on artificial intelligence words art model characterized by comprising
Entity extraction unit inputs the name entity in information for keyword extraction user;
Model Matching unit, for determining position of the name entity in words art model;
Knowledge extraction unit, for extracting name entity in the various knowledge informations of words art modal position information;
User is inputted the knowledge information phase of the feature of problem and name entity in information by deep learning by information answer unit
Association, determines final response content.
7. the matching system according to claim 6 based on artificial intelligence words art model, it is characterised in that: the words art mould
Type is a kind of database based on figure, and the content of storage is name entity, is associated between each name entity by attribute.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810825399.5A CN109062896A (en) | 2018-07-25 | 2018-07-25 | A kind of matching process and system based on artificial intelligence words art model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810825399.5A CN109062896A (en) | 2018-07-25 | 2018-07-25 | A kind of matching process and system based on artificial intelligence words art model |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109062896A true CN109062896A (en) | 2018-12-21 |
Family
ID=64836008
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810825399.5A Pending CN109062896A (en) | 2018-07-25 | 2018-07-25 | A kind of matching process and system based on artificial intelligence words art model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109062896A (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140280114A1 (en) * | 2013-03-15 | 2014-09-18 | Google Inc. | Question answering using entity references in unstructured data |
CN105095195A (en) * | 2015-07-03 | 2015-11-25 | 北京京东尚科信息技术有限公司 | Method and system for human-machine questioning and answering based on knowledge graph |
-
2018
- 2018-07-25 CN CN201810825399.5A patent/CN109062896A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140280114A1 (en) * | 2013-03-15 | 2014-09-18 | Google Inc. | Question answering using entity references in unstructured data |
CN105095195A (en) * | 2015-07-03 | 2015-11-25 | 北京京东尚科信息技术有限公司 | Method and system for human-machine questioning and answering based on knowledge graph |
Non-Patent Citations (1)
Title |
---|
周博通: "基于知识库的自动问答关键技术研究", 《中国优秀博硕士学位论文全文数据(硕士)信息科技辑(月刊)》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107291783B (en) | Semantic matching method and intelligent equipment | |
JP6793975B2 (en) | Video-based Jobs Job Matching Servers and methods and computer-readable recording media containing programs to perform those methods | |
CA2929018C (en) | Natural expression processing method, processing and response method, device and system | |
US8818926B2 (en) | Method for personalizing chat bots | |
US9081411B2 (en) | Rapid development of virtual personal assistant applications | |
CN110110041A (en) | Wrong word correcting method, device, computer installation and storage medium | |
CN108804698A (en) | Man-machine interaction method, system, medium based on personage IP and equipment | |
CN106897559B (en) | A kind of symptom and sign class entity recognition method and device towards multi-data source | |
CN111708869B (en) | Processing method and device for man-machine conversation | |
CN109101545A (en) | Natural language processing method, apparatus, equipment and medium based on human-computer interaction | |
US20200075024A1 (en) | Response method and apparatus thereof | |
CN108847241A (en) | It is method, electronic equipment and the storage medium of text by meeting speech recognition | |
CN108682420A (en) | A kind of voice and video telephone accent recognition method and terminal device | |
O’Shea et al. | Systems engineering and conversational agents | |
CN111767385A (en) | Intelligent question and answer method and device | |
CN111310440A (en) | Text error correction method, device and system | |
CN109902187A (en) | A kind of construction method and device, terminal device of feature knowledge map | |
CN107885703A (en) | Mathematics translater, mathematics interpreting equipment and platform | |
CN106776808A (en) | Information data offering method and device based on artificial intelligence | |
CN116401354A (en) | Text processing method, device, storage medium and equipment | |
CN111553138B (en) | Auxiliary writing method and device for standardizing content structure document | |
CN116187320A (en) | Training method and related device for intention recognition model | |
CN109190116A (en) | Semantic analytic method, system, electronic equipment and storage medium | |
CN116702736A (en) | Safe call generation method and device, electronic equipment and storage medium | |
CN110491372A (en) | A kind of feedback information generating method, device, storage medium and smart machine |
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: 20181221 |
|
RJ01 | Rejection of invention patent application after publication |