Research On Semantic Web Service Matching Algorithm Based On Similarity Of Ontology

With the development of global network and informationization, network servicesare growing rapidly. Then here comes a hot issue, that is, how to use network servicesquickly and efficiently to meet your needs. The development of semantic Webtechnology has opened a window for us.Traditional Web services are implemented by keyword similarity matching. Butdue to the insufficient semantic information, Web services, to a certain extent, havedefects. With the introduction of the concept of ontology and the semantic Web servicesdescription language (WSDL), the expression and discovery ability of semantics havebeen greatly improved. Thus, the computer is able to not only understand and processsemantic information to the maximum, but also support automatical discovery. And itcan improve the recall rate and precision. The main aim of this thesis is to discuss howto combine the keyword research and Web services description perfectly and explorethe similarity calculation and semantic Web service matching accuracy.The thesis forwaods a new hybrid calculation method based on the conceptdifferences of semantic Web ontology, by comparing the two kinds of semanticsimilarity calculation methods of network distance model and information model and bythe analysis of the advantages and disadvantages of two algorithms. The methodcombines advantages of the two algorithms of network distance model and informationmodel and avoid its defects and disadvantages to the maximum. All aspects ofinterference factors are considered in order to reach the ideal goal. Furthermore, theways of semantic reasoning and text abstracting are further studied in order to make theissue accurate as far as possible to get the minimum range of semantic feature sets andto perfect matching effect. Finally, making use of the concept correlation feature and itsasymmetry, we undertake matrix conversion by using the annotated concepts ofsemantic feature, convert it into SFS concept of semantic feature set and model FFS,and implement the final service match by calculating the similarity.