CN106960039B - Social search engine system based on whole multi-Agent - Google Patents
Social search engine system based on whole multi-Agent Download PDFInfo
- Publication number
- CN106960039B CN106960039B CN201710190053.8A CN201710190053A CN106960039B CN 106960039 B CN106960039 B CN 106960039B CN 201710190053 A CN201710190053 A CN 201710190053A CN 106960039 B CN106960039 B CN 106960039B
- Authority
- CN
- China
- Prior art keywords
- module
- agent
- holon
- data
- user
- 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
- 238000005065 mining Methods 0.000 claims abstract description 73
- 238000004140 cleaning Methods 0.000 claims abstract description 48
- 238000005457 optimization Methods 0.000 claims abstract description 46
- 239000003795 chemical substances by application Substances 0.000 claims description 173
- 238000004891 communication Methods 0.000 claims description 64
- 238000004458 analytical method Methods 0.000 claims description 30
- 239000012459 cleaning agent Substances 0.000 claims description 13
- 238000012544 monitoring process Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000012163 sequencing technique Methods 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 7
- 238000012423 maintenance Methods 0.000 claims description 2
- 239000011159 matrix material Substances 0.000 claims description 2
- 239000003599 detergent Substances 0.000 claims 1
- 238000005516 engineering process Methods 0.000 abstract description 7
- 206010063385 Intellectualisation Diseases 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 17
- 238000000034 method Methods 0.000 description 10
- 230000008569 process Effects 0.000 description 4
- 239000013598 vector Substances 0.000 description 4
- 230000007547 defect Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000002372 labelling Methods 0.000 description 2
- 238000013481 data capture Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2216/00—Indexing scheme relating to additional aspects of information retrieval not explicitly covered by G06F16/00 and subgroups
- G06F2216/03—Data mining
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Computing Systems (AREA)
- Human Resources & Organizations (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Fuzzy Systems (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a social search engine system based on a whole-son multi-Agent, which comprises a data acquisition and cleaning father Holon module, a knowledge mining father Holon module, a search optimization father Holon module, a search cache library, a user information library and a user knowledge library. Acquiring and cleaning data by a data acquisition and cleaning father Holon module; a knowledge mining parent Holon module mines social knowledge implied in social data; the search optimization parent Holon module provides personalized query results and an intelligent information recommendation function for the user. The social search engine system is designed by adopting the whole multi-Agent technology and fully utilizing the characteristics of perceptibility, collaboration, expandability and the like of the whole multi-Agent technology, and has the advantages of high degree of individuation and intellectualization and good expandability.
Description
Technical Field
the invention belongs to the technical field of communication, and further relates to a social search engine system based on a whole multi-Agent in the technical field of information retrieval. The invention is mainly used for optimizing the retrieval result of the full-text search engine and providing diversified information recommendation functions for the user, so that the retrieval process becomes more personalized and intelligent.
Background
Currently, with the rapid development of internet technology, search engines become important tools for information retrieval on the internet. Social search (social search) aims to make up for the defects of a traditional search engine by using the advantages of a social network, attaches importance to the initiative of group collaboration in information retrieval, optimizes search results through mutual connection among users, and provides high-quality and high-relevance search service. However, the current social search engine has the following three problems: the personalized service level is low due to lack of deep mining on multi-dimensional knowledge of user interest, social relations and influence; the existing social search engine system is lack of perceptibility, collaboration and expandability, so that the system intelligence degree is low; there is a lack of a function of recommending contents of interest to the user.
The patent document of university of college studios' a personalized search method and system based on social annotation (patent application No. 2015102465031, publication No. CN 104866554A) discloses a method and system for optimizing search results by using social annotation. The system comprises a webpage document processing module, a related vector extraction module, a user similarity calculation module, a similar user selection module, a calculation module of personalized tag vectors of a user to documents, an extended attribute vector calculation module of the user and a document typing and sorting module. According to the system, the interest preference of the user is deeply mined by using the tag information disclosed by the user, namely, personalized search optimization is carried out by using the information actively disclosed by the user from the labeled information of the webpage user, so that the problems of privacy disclosure and cold start are avoided, and the retrieval accuracy is better improved. In addition, the extended attribute vector calculation module of the user searches for the high-similarity user according to the webpage category information and the label information browsed by the user, and forms the extended attribute of the target user by using the information of the similar users, so that the interest description of the user is more comprehensive and effective. However, the system still has the following disadvantages: firstly, comparing one surface of user interest mined completely based on labeling information; secondly, the similar user selection module endows all users with the same similarity value with the same weight, and a multidimensional personalized optimization strategy is lacked; finally, the system lacks intelligent recommendation functionality and is unable to recommend content or users to a user that may be of interest.
a socialized search engine method and system (patent application No. 2012104411846, publication No. CN 102930029A) disclosed in the patent document applied by Beijing network Intelligent Tianyuan science and technology Co Ltd is a socialized search method and system for optimizing search results by using microblog user data. The system consists of an information capturing module, an information extracting module, an expert database and an inquiry request processing module. The information capture module is responsible for extracting the basic information of the user from the microblog and establishing an expert information base. Then the query request processing module finds the experts related to the query words in the expert information base according to the query words input by the user, forwards the query words to one or more selected experts according to the autonomous selection of the user, tracks the request in real time, timely captures the results returned by the experts, processes the query results and returns the results to the current user. The system considers human factors more, and realizes a human-to-human information acquisition mode by helping a user to find an expert most relevant to the query word, so that the recall ratio and precision ratio of the search engine are improved. The system still has the following defects: firstly, when the influence of the expert is considered, only the attention number, the fan number and the issued Bowen number of the user are considered, but the influence of the comment number and the forwarding number on the influence of the expert is ignored; secondly, when searching for relevant experts, sorting is carried out only depending on the expertise of the experts, but the social association between the current user and the experts is ignored; finally, the system can only perform personalized optimization aiming at the search results of the users in the social network, but cannot optimize the results of the full-text search engine, and the system lacks the characteristics of expandability, autonomy, perceptibility, intellectualization and the like.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a social search engine system based on a whole multi-Agent. The method and the system deeply analyze the user interests, the community relations and the influence by utilizing the social network data of the user, optimize the retrieval results returned by the full-text search engine based on the social knowledge, and recommend the interested information to the user. In addition, the whole multi-Agent system architecture can also improve the perceptibility, the collaboration and the expandability of the social search engine system, and provide more personalized and intelligent retrieval experience for users.
The specific thought of the invention is as follows: the social search engine system is realized by using an intelligent Agent technology, an organization mode among intelligent agents is designed based on a whole structure, the mining of user interest, the division of community relations and the evaluation of user influence are realized by using the characteristics of perception, collaboration, expandability and the like of a plurality of agents of the whole structure, the search results of a full-text search engine are reordered, and diversified information recommendation is provided for users.
In order to achieve the purpose, the social search engine system based on the whole-child multi-Agent provided by the invention comprises a data acquisition and cleaning father Holon module, a user information base, a search cache base, a knowledge mining father Holon module, a user knowledge base and a search optimization father Holon module, wherein the functions of the modules are as follows:
the data acquisition and cleaning parent Holon module comprises a data center Agent and a plurality of data acquisition and cleaning child Holon modules. The data center Agent is used for generating the data acquisition and cleaning sub-Holon module when resources are in short according to the load information of the data acquisition and cleaning sub-Holon module, and destroying redundant data acquisition and cleaning sub-Holon module when resources are excessive so as to realize load balance. And the data acquisition and cleaning sub-Holon module is used for storing search engine retrieval result data into a search cache library and storing cleaned social network data into a user information library.
And the user information base is used for storing the social network data obtained by data acquisition and washed by the parent Holon module and providing data required by knowledge mining for the knowledge mining parent Holon module.
And the search cache library is used for storing the retrieval result of the full-text search engine tool and is used as a data source of the retrieval result required by the search optimization parent Holon module.
The knowledge mining father Holon module comprises a mining center Agent and a plurality of knowledge mining son Holon modules. And the mining center Agent is used for generating the knowledge mining sub-Holon module when the resources are in short according to the load information of the knowledge mining sub-Holon module, and destroying the redundant knowledge mining sub-Holon module when the resources are excessive so as to realize load balance. The knowledge mining sub-Holon module is used for mining three types of social knowledge including user interest knowledge, community knowledge and influence knowledge which are hidden in social network data, and storing the obtained social knowledge in a user knowledge base.
And the user knowledge base is used for storing the social knowledge mined by the knowledge mining father Holon module and serving as a knowledge source required by the search optimization father Holon module for optimizing the search result.
the search optimization parent Holon module comprises an optimization center Agent and a plurality of search optimization child Holon modules. The optimization center Agent is used for generating the search optimization sub-Holon module when the resources are in short according to the load information of the search optimization sub-Holon module, and destroying the redundant search optimization sub-Holon module when the resources are excessive so as to realize load balance. And the search optimization sub-Holon module is used for optimizing the retrieval result in the search cache library by utilizing social knowledge in the user knowledge base and recommending interested information to the user.
Compared with the prior art, the invention has the following advantages:
Firstly, the knowledge mining father Holon module reasonably utilizes online social network data, deeply mines user interest knowledge, community knowledge and influence knowledge by mining the social network data offline, and solves the problem of low personalized service level caused by the fact that the user interest is obtained only based on user labeling or browsing history records in the prior art, so that the invention can accurately return more personalized retrieval results according to the user interest.
Secondly, the social search engine system is designed by adopting a whole multi-Agent technology and fully utilizing the characteristics of perceptibility, collaboration, expandability and the like of the whole multi-Agent technology, and the problem of low system intelligence degree caused by lack of perceptibility, collaboration and expandability in the conventional social search engine system is solved. The method and the device can actively sense the social knowledge change of the user in the operation process, update the social knowledge of the user in real time, enable the user to quickly find the required information, and improve the perceptibility, the collaboration and the expandability of the social search engine.
Third, the search optimization father-Holon module utilizes knowledge to mine social knowledge mined by the father-Holon module, so that the recommendation function of relevant query words and community users based on the social knowledge is realized, the problem that information recommendation of users in interest is lacked in the prior art is solved, the content of interest of the current users is analyzed based on group intelligence in the operation process, and the search experience of the users is improved.
Drawings
FIG. 1 is a schematic structural view of the present invention;
FIG. 2 is a schematic diagram of the data acquisition and cleaning parent Holon module of the present invention;
FIG. 3 is a schematic structural diagram of a knowledge mining parent Holon module of the present invention;
Fig. 4 is a schematic structural diagram of a search optimization parent Holon module according to the present invention.
Detailed Description
the present invention is described in further detail below with reference to the attached drawing figures.
referring to fig. 1, the present invention includes a data acquisition and cleaning parent Holon module, a user information base, a search cache base, a knowledge mining parent Holon module, a user knowledge base, and a search optimization parent Holon module. According to the invention, through mutual cooperation of the three types of Holon modules, the user interest, the user community relation and the user influence are fully mined by utilizing social network data, the retrieval result of a full-text search engine tool is optimized, diversified information recommendation is provided, and finally, a more personalized and intelligent social search engine system is realized. The data acquisition and cleaning parent Holon module comprises a data center Agent and a plurality of data acquisition and cleaning child Holon modules. The data center Agent is used for dynamically generating or destroying the data acquisition and cleaning son Holon module during load balancing. And the data acquisition and cleaning sub Holon module is used for acquiring and cleaning data. And the user information base is used for storing the social network data. And the search cache library is used for storing the retrieval result of the full-text search engine tool. The knowledge mining father Holon module comprises a mining center Agent and a plurality of knowledge mining son Holon modules. And the mining center Agent is used for dynamically generating or destroying the knowledge mining sub Holon module during load balancing. The knowledge mining sub-Holon module is used for mining social knowledge of the user and storing the acquired social knowledge into a user knowledge base. And the user knowledge base is used for storing the social knowledge mined by the knowledge mining father Holon module. The search optimization parent Holon module comprises an optimization center Agent and a plurality of search optimization child Holon modules. And the optimization center Agent is used for dynamically generating or destroying the search optimization sub-Holon module when realizing load balancing. And the search optimization sub-Holon module is used for optimizing the retrieval result in the search cache library and recommending the information of interest to the user.
referring to fig. 2, the data center Agent includes a holoon manager, a load status library, a load monitor, and a communication module. And the Holon manager is used for reading the load information of the data acquisition and cleaning sub-Holon module from the load state library, generating the data acquisition and cleaning sub-Holon module when the resources are in short, and destroying the redundant data acquisition and cleaning sub-Holon module when the resources are excessive. And the load state library is used for storing the task load information of the data acquisition and cleaning sub-Holon module. And the load monitor is used for monitoring the task load condition of the data acquisition and cleaning sub-Holon module. And the communication module is used for realizing the information communication function between the data acquisition and cleaning parent Holon module and the data acquisition and cleaning child Holon module. The data acquisition and cleaning sub-Holon module comprises a data center sub-Agent and a plurality of data acquisition and cleaning agents. The data center sub-Agent comprises a load monitor, an Agent manager and a communication module. And the load monitor is used for monitoring the task load condition of the data acquisition and cleaning Agent. And the Agent manager is used for generating data acquisition and cleaning agents when the resources are in short according to the load condition sensed by the load monitor, and destroying redundant data acquisition and cleaning agents when the resources are excessive so as to realize load balance. And the communication module is used for realizing the information exchange functions of the data acquisition and cleaning son Holon module and the data acquisition and cleaning Agent. The data acquisition and cleaning Agent comprises a communication module, a data type sensor, a data updating module, a data capturing module, a data cleaning module and a data cache library. And the communication module is used for realizing the information exchange function of the data acquisition and cleaning Agent and the data center sub-Agent. And the data type perceptron is used for perceiving the acquired data type and transmitting the perceived result to the data updating module and the data capturing module. The data updating module is used for receiving the result transmitted by the data type perceptron and storing the search engine retrieval result data into a search cache library; and the data updating module receives the social network data cleaned by the data cleaning module and stores the social network data in the user information base. And the data capturing module is used for receiving the result transmitted by the data type perceptron and storing the social network data into the data cache library. And the data cleaning module is used for cleaning the social network data in the data cache library and submitting the cleaned data to the data updating module. And the data cache library is used for storing the data transmitted by the data capture module.
Referring to fig. 3, the mining center Agent includes a holoon manager, a load status library, a load monitor, and a communication module. And the Holon manager is used for reading the load information of the knowledge mining sub-Holon module from the load state library, generating the knowledge mining sub-Holon module when the resources are in short, and destroying the redundant knowledge mining sub-Holon module when the resources are excessive. And the load state library is used for storing task load information of the knowledge mining sub-Holon module. And the load monitor is used for monitoring the task load condition of the knowledge mining sub-Holon module. And the communication module is used for realizing the information exchange function between the knowledge mining father Holon module and the knowledge mining son Holon module. The knowledge mining sub-Holon module comprises a mining center sub-Agent, a user Agent, an interest mining Agent, a community dividing Agent and an influence analysis Agent. The mining center sub-Agent comprises a load monitor, an Agent manager and a communication module. And the load monitor is used for monitoring the task load conditions of the interest mining Agent, the community dividing Agent and the influence analysis Agent. And the Agent manager generates each Agent when the resource is in shortage according to the load condition sensed by the load monitor, and destroys each Agent when the resource is in excess so as to realize load balance. And the communication module is used for realizing the information exchange function of the knowledge mining sub-Holon module and the interest mining Agent, the community dividing Agent and the influence analysis Agent. The user Agent comprises a communication module, a data change perceptron and a data updating module. And the communication module is used for realizing the information exchange function of the user Agent, the interest mining Agent, the community dividing Agent and the influence analysis Agent. And the data change perceptron is used for perceiving data change in the user information base and informing the interest mining Agent, the influence analysis Agent and the community dividing Agent in time. And the data updating module is used for updating the user knowledge base. The interest mining Agent comprises a communication module and an interest inference machine module. And the communication module is used for realizing the information exchange function of the interest mining Agent, the user Agent, the community dividing Agent and the influence analysis Agent. And the interest inference machine module analyzes the social network data in the user information base by using a Bayesian formula to obtain the interest degree of each interest item of the user, so as to maintain and update the user interest, and delivers the result to the user Agent through the communication module. The community dividing Agent comprises a communication module, a user similarity calculation module, a user similarity knowledge base, a user similarity community inference machine and a user interest community inference machine. And the communication module is used for realizing the information exchange function of the community dividing Agent, the user Agent, the interest mining Agent and the influence analysis Agent. And the user similarity calculation module is used for calculating the similarity between the users by using the basic information of the users in the social network data and transmitting the result to the user similarity knowledge base. And the user similarity knowledge base is used for storing the data transmitted by the similarity calculation module. And the user similarity community inference machine divides the user communities by utilizing the data in the similarity knowledge base. And the user interest community inference machine utilizes the analysis result of the interest mining Agent to realize the user community relation division from the user interest dimension and delivers the result to the user Agent through the communication module. The influence analysis Agent comprises a communication module, a field classifier module and a fine-grained influence inference machine. And the communication module is used for realizing the information exchange function of the influence analysis Agent, the user Agent, the interest mining Agent and the community dividing Agent. And the domain classifier module is used for identifying all domain information to which the user belongs by utilizing the social network data in the user information base. And the fine-grained influence inference machine is used for mining the influence values of the users in different fields, realizing the maintenance and the update of the influence of the users and delivering the result to the user Agent through the communication module.
Referring to fig. 4, the optimization center Agent includes a holoon manager, a load status library, a load monitor, and a communication module. And the Holon manager is used for reading the load information of the search optimization sub-Holon module from the load state library, generating the search optimization sub-Holon module when the resources are in short, and destroying the redundant search optimization sub-Holon module when the resources are excessive. And the load state library is used for storing task load information of the search optimization sub-Holon module. And the load monitor is used for monitoring the task load condition of the search optimization sub-Holon module. And the communication module is used for realizing the information communication function between the search optimization parent Holon module and the search optimization child Holon module. The search optimization sub-Holon module comprises an optimization center sub-Agent, a result sorting Agent and an information recommending Agent. The optimization center sub-Agent comprises a load monitor, an Agent manager and a communication module. And the load monitor is used for monitoring the task load conditions of the result sequencing Agent and the information recommending Agent. And the Agent manager generates each Agent when the resource is in shortage according to the load condition sensed by the load monitor, and destroys each Agent when the resource is in excess so as to realize load balance. And the communication module is used for realizing the information exchange function of the search optimization sub-Holon module, the result sequencing Agent and the information recommendation Agent. The result sequencing Agent comprises a communication module, a retrieval result analyzer, a result analysis knowledge base and a result sequencer. And the communication module is used for realizing the information exchange function of the optimization center sub-Agent and the result sequencing Agent. And the retrieval result analyzer is used for analyzing the retrieval results in the search cache library and delivering the analysis results to the result analysis knowledge base. And the result analysis knowledge base is used for storing the data transmitted by the retrieval result analyzer. And the result sequencer is used for utilizing the social knowledge in the user knowledge base to carry out personalized reordering on the retrieval results in the result analysis knowledge base. The information recommending Agent comprises a communication module, a community user recommender module and a related query word recommender module. And the communication module is used for realizing the information exchange function of the optimization center sub-Agent and the information recommendation Agent. And the community user recommender module is used for sorting the users in the community in a descending order according to the influence value and recommending the top 5 users to the users. And the related query word recommender module is used for comparing the similarity between the query words of the recommended users generated by the community user recommender module in about 3 days and the query words input by the current users to obtain a similarity matrix, and recommending the query words with the similarity values arranged in the top 5 names to the current users.
Claims (7)
1. a socialized search engine system based on a whole-child multi-Agent comprises a data acquisition and cleaning father Holon module, a user information base, a search cache base, a knowledge mining father Holon module, a user knowledge base and a search optimization father Holon module, wherein:
The data acquisition and cleaning parent Holon module comprises a data center Agent and a plurality of data acquisition and cleaning child Holon modules; the data center Agent is used for dynamically generating or destroying the data acquisition and cleaning son Holon module during load balancing; the data acquisition and cleaning sub-Holon module is used for storing search engine retrieval result data into a search cache library and storing cleaned social network data into a user information library;
The user information base is used for storing the social network data obtained after the data acquisition and cleaning of the parent Holon module is cleaned, and providing data required by knowledge mining for the knowledge mining parent Holon module;
the search cache library is used for storing the retrieval result of the full-text search engine tool and is used as a data source of the retrieval result required by the search optimization father Holon module;
The knowledge mining father Holon module comprises a mining center Agent and a plurality of knowledge mining son Holon modules; the mining center Agent is used for dynamically generating or destroying a knowledge miner Holon module when load balancing is realized; the knowledge mining sub-Holon module is used for analyzing social network data stored in a user information base, mining three types of social knowledge including interest knowledge, community knowledge and influence knowledge of a user, and storing the obtained social knowledge into the user knowledge base;
The user knowledge base is used for storing social knowledge mined by the knowledge mining father Holon module and used as a knowledge source required by the search optimization father Holon module for optimizing search results;
the search optimization parent Holon module comprises an optimization center Agent and a plurality of search optimization child Holon modules; the optimization center Agent is used for dynamically generating or destroying a search optimization sub-Holon module when load balancing is realized; the search optimization sub-Holon module is used for optimizing retrieval results in the search cache library and recommending interesting information to the user.
2. The social search engine system based on whole-son multi-Agent of claim 1, wherein the data center Agent comprises a holoon manager, a load status library, a load monitor and a communication module, wherein:
the Holon manager is used for reading load information of the data acquisition and cleaning sub-Holon module from the load state library, generating the data acquisition and cleaning sub-Holon module when resources are in short, and destroying redundant data acquisition and cleaning sub-Holon module when resources are excessive;
The load state library is used for storing task load information of the data acquisition and cleaning sub-Holon module;
the load monitor is used for monitoring the task load condition of the data acquisition and cleaning sub-Holon module;
the communication module is used for realizing the information communication function between the data acquisition and cleaning parent Holon module and the data acquisition and cleaning child Holon module.
3. The social search engine system based on integer multiple agents of claim 1, wherein the data acquisition and cleansing sub-Holon module comprises a data center sub-Agent and multiple data acquisition and cleansing agents, wherein:
the data center sub-Agent comprises a load monitor, an Agent manager and a communication module; the load monitor is used for monitoring the task load condition of the data acquisition and cleaning Agent; the Agent manager is used for generating data acquisition and cleaning agents when the resources are in short according to the load condition sensed by the load monitor, and destroying redundant data acquisition and cleaning agents when the resources are excessive so as to realize load balance; the communication module is used for realizing the information exchange function of the data acquisition and cleaning son Holon module and the data acquisition and cleaning Agent;
The data acquisition and cleaning Agent comprises a communication module, a data type sensor, a data updating module, a data capturing module, a data cleaning module and a data cache library; the communication module is used for realizing the information exchange function of the data acquisition and cleaning Agent and the data center sub-Agent; the data type sensor is used for sensing the acquired data type and transmitting the sensed result to the data updating module and the data capturing module; the data updating module receives the result transmitted by the data type perceptron and stores the search engine retrieval result data into a search cache library; the data updating module receives the social network data cleaned by the data cleaning module and stores the social network data in the user information base; the data capturing module receives a result transmitted by the data type perceptron and stores the social network data into a data cache library; the data cleaning module is used for cleaning social network data in the data cache library and submitting the cleaned data to the data updating module; the data cache bank is used for storing the data transmitted by the data capturing module.
4. The social search engine system based on whole multi-Agent of claim 1, wherein the mining center Agent comprises a holoon manager, a load status library, a load monitor and a communication module, wherein:
the Holon manager is used for reading the load information of the knowledge mining sub-Holon module from the load state library, generating the knowledge mining sub-Holon module when the resources are in short, and destroying the redundant knowledge mining sub-Holon module when the resources are excessive;
The load state library is used for storing task load information of the knowledge mining sub-Holon module;
the load monitor is used for monitoring the task load condition of the knowledge mining sub-Holon module;
The communication module is used for realizing the information communication function between the knowledge mining father Holon module and the knowledge mining son Holon module.
5. The social search engine system based on integer multiple agents of claim 1, wherein the knowledge mining sub-Holon module comprises a mining center sub-Agent, an interest mining Agent, a community dividing Agent, an influence analysis Agent and a user Agent, wherein:
The mining center sub-Agent comprises a load monitor, an Agent manager and a communication module; the load monitor is used for monitoring the task load conditions of the interest mining Agent, the community dividing Agent and the influence analysis Agent; the Agent manager generates each Agent when the resource is in shortage according to the load condition sensed by the load monitor, and destroys each Agent when the resource is excessive so as to realize load balance; the communication module is used for realizing the information exchange function of the knowledge mining sub-Holon module and the interest mining Agent, the community dividing Agent and the influence analysis Agent;
the interest mining Agent comprises a communication module and an interest inference machine module; the communication module is used for realizing the information exchange function of the interest mining Agent, the user Agent, the community dividing Agent and the influence analysis Agent; the interest inference engine module analyzes social network data in the user information base by using a Bayesian formula to obtain the interest degree of each interest item of the user, so as to realize maintenance and update of the user interest, and delivers the result to the user Agent through the communication module;
the community dividing Agent comprises a communication module, a user similarity calculation module, a user similarity knowledge base, a user similarity community inference machine and a user interest community inference machine; the communication module is used for realizing the information exchange function of the community dividing Agent, the user Agent, the interest mining Agent and the influence analysis Agent; the user similarity calculation module is used for calculating the similarity between users by using the basic information of the users in the social network data and transmitting the result to the user similarity knowledge base; the user similarity knowledge base is used for storing the data transmitted by the similarity calculation module; the user similarity community inference machine divides the user communities by utilizing data in the similarity knowledge base; the user interest community inference machine utilizes the analysis result of the interest mining Agent to realize the user community relation division from the user interest dimension and delivers the result to the user Agent through the communication module;
The influence analysis Agent comprises a communication module, a field classifier module and a fine-grained influence inference machine; the communication module is used for realizing the information exchange function of the influence analysis Agent, the user Agent, the interest mining Agent and the community dividing Agent; the domain classifier module is used for identifying all domain information to which the user belongs by utilizing social network data in the user information base; the fine-grained influence inference machine is used for mining influence values of users in different fields, maintaining and updating the influence of the users and delivering results to the user agents through the communication module;
the user Agent comprises a communication module, a data change perceptron and a data updating module; the communication module is used for realizing the information exchange function of the user Agent, the interest mining Agent, the community dividing Agent and the influence analysis Agent; the data change sensor is used for sensing data change in the user information base and timely notifying an interest mining Agent, an influence analysis Agent and a community dividing Agent; and the data updating module is used for updating the user knowledge base.
6. the social search engine system based on whole-son multi-Agent of claim 1, wherein the optimization center Agent comprises a holoon manager, a load status library, a load monitor and a communication module, wherein:
The Holon manager is used for reading load information of the search optimization sub-Holon module from the load state library, generating the search optimization sub-Holon module when resources are in short, and destroying redundant search optimization sub-Holon modules when resources are excessive;
the load state library is used for storing task load information of the search optimization sub-Holon module;
The load monitor is used for monitoring the task load condition of the search optimization sub-Holon module;
the communication module is used for realizing the information communication function between the search optimization parent Holon module and the search optimization child Holon module.
7. The social search engine system based on whole-son multi-Agent of claim 1, wherein the search optimization sub-Holon module comprises an optimization center sub-Agent, a result ranking Agent and an information recommendation Agent, wherein:
The optimization center sub-Agent comprises a load monitor, an Agent manager and a communication module; the load monitor is used for monitoring the task load conditions of the result sequencing Agent and the information recommending Agent; the Agent manager generates each Agent when the resource is in shortage according to the load condition sensed by the load monitor, and destroys each Agent when the resource is excessive so as to realize load balance; the communication module is used for realizing the information exchange function of the search optimization sub-Holon module, the result sequencing Agent and the information recommendation Agent;
The result sequencing Agent comprises a communication module, a retrieval result analyzer, a result analysis knowledge base and a result sequencer; the communication module is used for realizing the information exchange function of the optimization center sub-Agent and the result sequencing Agent; the retrieval result analyzer is used for analyzing retrieval results in the search cache library and delivering the analysis results to the result analysis knowledge base; the result analysis knowledge base is used for storing data transmitted by the retrieval result analyzer; the result sequencer utilizes social knowledge in the user knowledge base to carry out personalized reordering on the retrieval results in the result analysis knowledge base;
The information recommending Agent comprises a communication module, a community user recommender module and a related query word recommender module; the communication module is used for realizing the information exchange function of the optimization center sub-Agent and the information recommendation Agent; the community user recommender module is used for sorting the users in the community in a descending order according to the influence value and recommending the top 5 users to the users; the related query word recommender module is used for comparing the similarity between the query words of the recommended users generated by the community user recommender module in about 3 days and the query words input by the current users to obtain a similarity matrix, and recommending the query words with the similarity values arranged in the top 5 names to the current users.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710190053.8A CN106960039B (en) | 2017-03-28 | 2017-03-28 | Social search engine system based on whole multi-Agent |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710190053.8A CN106960039B (en) | 2017-03-28 | 2017-03-28 | Social search engine system based on whole multi-Agent |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106960039A CN106960039A (en) | 2017-07-18 |
| CN106960039B true CN106960039B (en) | 2019-12-10 |
Family
ID=59471185
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710190053.8A Active CN106960039B (en) | 2017-03-28 | 2017-03-28 | Social search engine system based on whole multi-Agent |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106960039B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110609834B (en) * | 2018-05-29 | 2023-04-18 | 西安电子科技大学 | Multi-source heterogeneous government affair data extraction system based on Agent |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012138611A2 (en) * | 2011-04-08 | 2012-10-11 | Zalila-Wenkstern Rym | Customizable interfacing agents, systems, and methods |
| CN103034672A (en) * | 2011-09-29 | 2013-04-10 | 云壤（北京）信息技术有限公司 | Social search system and social search method |
| CN104462373A (en) * | 2014-12-09 | 2015-03-25 | 南京大学 | Personalized recommendation engine implementing method based on multiple Agents |
| CN102902800B (en) * | 2012-10-12 | 2015-06-03 | 西安电子科技大学 | Agent-based intelligent meta search engine system |
| CN104866554A (en) * | 2015-05-15 | 2015-08-26 | 大连理工大学 | Personalized searching method and system on basis of social annotation |
-
2017
- 2017-03-28 CN CN201710190053.8A patent/CN106960039B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012138611A2 (en) * | 2011-04-08 | 2012-10-11 | Zalila-Wenkstern Rym | Customizable interfacing agents, systems, and methods |
| CN103034672A (en) * | 2011-09-29 | 2013-04-10 | 云壤（北京）信息技术有限公司 | Social search system and social search method |
| CN102902800B (en) * | 2012-10-12 | 2015-06-03 | 西安电子科技大学 | Agent-based intelligent meta search engine system |
| CN104462373A (en) * | 2014-12-09 | 2015-03-25 | 南京大学 | Personalized recommendation engine implementing method based on multiple Agents |
| CN104866554A (en) * | 2015-05-15 | 2015-08-26 | 大连理工大学 | Personalized searching method and system on basis of social annotation |
Non-Patent Citations (3)
| Title |
|---|
| Improving Search In Social Networks by Agent Based Mining;Anıl G¨ursel 等;《Proceedings of the Twenty-First International Joint Conference on Artificial Intelligence》;20091231;2034-2039 * |
| The impactofdiversityonperformanceofholonicmulti-agentsystems;Ahmad Esmaeili 等;《Engineering ApplicationsofArtificial Intelligence55》;20161231;186–201 * |
| 社交网络相关机制应用于搜索引擎的研究综述;费洪晓 等;《计算技术与自动化》;20140331;第33卷(第1期);1-9 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106960039A (en) | 2017-07-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chu et al. | A hybrid recommendation system considering visual information for predicting favorite restaurants | |
| Aligon et al. | A collaborative filtering approach for recommending OLAP sessions | |
| CN105512183B (en) | A kind of personalized recommendation method independently selected based on user and system | |
| US9536003B2 (en) | Method and system for hybrid information query | |
| CN103324657B (en) | Extended using the multi-dimensional query of semantic and service condition statistics | |
| US20180239829A1 (en) | Providing query explanations for automated sourcing | |
| US10600011B2 (en) | Methods and systems for improving engagement with a recommendation engine that recommends items, peers, and services | |
| CN104933100A (en) | Keyword recommendation method and device | |
| CA2675216A1 (en) | Method and system for information discovery and text analysis | |
| CN103646092A (en) | SE (search engine) ordering method based on user participation | |
| Wu et al. | An incremental community detection method for social tagging systems using locality-sensitive hashing | |
| CN116401459A (en) | Internet information processing method, system and recording medium | |
| CN115130008A (en) | Search ordering method based on machine learning model algorithm | |
| CN104484367A (en) | Data mining and analyzing system | |
| CN106960039B (en) | Social search engine system based on whole multi-Agent | |
| Bhujbal et al. | News aggregation using web scraping news portals | |
| CN117540087A (en) | Personalized news data recommendation method, system and storage medium | |
| Bhushan et al. | Automatic recommendation of web pages for online users using web usage mining | |
| Bimonte et al. | Evaluation of user satisfaction with OLAP recommender systems: an application to RecoOLAP on a agricultural energetic consumption datawarehouse | |
| CN115563189A (en) | Mass data query method based on data mining technology | |
| Thakur et al. | Query based personalization in semantic web mining | |
| Veningston et al. | Enhancing personalized web search re-ranking algorithm by incorporating user profile | |
| Cai et al. | Learning search popularity for personalized query completion in information retrieval | |
| Faqeeh et al. | Topical search engine for Internet of Things | |
| Khelghati | Deep web content monitoring |
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 |