Research On WEB Semantic Structure Generation Based On Knowledge Graph

The semantic Web tries to give semantics to resources on the Internet so that computers can understand what resources really mean.However,the meaning of resources is not exactly the same in the eyes of the owner and user,which is the semantic self-consciousness.The meaning of each resource is determined by the meaning of the resource related to it,which is called semantic mutuality.The degree of correlation between resources is different,which is called semantic fuzziness.The relationship and degree of correlation between resources is constantly changing,which is called semantic dynamics.The semantic Web USES self-consciousness and mutual representation,but does not explicitly put forward these two concepts,and does not involve dynamics and ambiguity.The research team of explicitly puts forward these four characteristics of semantics and applies them to the design of semantic Web,and designs a semantic Web framework based on self-consciousness.Semantic Web is a based on self-awareness has website Agent,the Agent of mesh based on self-consciousness,and customer Agent of three kinds of Agent,including mesh Agent is a customer by customer Agent access by the window of the Web site of the Agent of network resources,is the core of the semantic Web based on self-awareness,mesh Agent's belief that mesh Agent understanding of the Internet,known as Web semantic structure.Although the semantic structure of the WEB will change with the user's access and the change of network resources,the generation of its initial structure is prepared by the knowledge map of the existing network resources.This paper mainly conducts the following research work on the generation of WEB semantic structure based on the knowledge map:1.The similarities and differences between knowledge graph and WEB semantic structure are studied.Knowledge graph is a structured semantic knowledge base.The WEB semantic structure follows the representation of knowledge in the ontology layer of knowledge graph,and USES the entity-relation-entity,entity-attribute-attribute value triples to represent knowledge,thus realizing the mutual representation of knowledge.2.The fuzziness of WEB semantic structure are studied,an improved fuzzy c-means clustering algorithm,according to the sample density distribution and the distance between the samples of the initial clustering center,and the class spacing factor into the objective function,so as to make the clustering results appear to avoid the local minimum and improve the clustering accuracy.3.The process of transforming knowledge map into WEB semantic structure is studied.To obtain entity classes in the knowledge map,according to the constructed fuzzy ontology,theimproved fuzzy clustering algorithm is used for fuzzy classification,and then the relationship class is established for the obtained fuzzy entity classes to generate the object part of the WEB semantic structure.In this paper,the set of acquaintances in WEB semantic structure is designed,and descriptive knowledge is adopted to represent the main body of acquaintances through four attributes of belief,ability,desire and planning,so as to complete the knowledge representation of the set of acquaintances in WEB semantic structure.4.The film field,select to implement WEB semantic structure,obtain film field data in structured data sources,using the top-down approach movie domain knowledge map construction,based on the field of film fuzzy ontology,in accordance with the rules of fuzzy ontology using the improved fuzzy c-means algorithm of knowledge map entity class data,fuzzy entity class,and to establish the relationship between each entity class class,generated movie field WEB object part of the semantic structure;Then by obtaining the movie website information to build the main body of the website acquaintances,and generate the acquaintance collection of the movie domain WEB semantic structure.5.Designed and implemented WEB semantic structure visualization program.To achieve data management,query display,manual adjustment and other functions,the generated movie domain WEB semantic structure as a data source,through the graphical way to display to the user.