Research On Key Issues Of Web Service Society Modeling Based On Multidimensional Semantic Awareness

Currently, Web service on the Internet is not only a huge number, but alsopresents the features of diversity and complexity, which makes how to precisely findthe requested services and organize them into complex business processes orcomposite services has become a critical issue for Service Oriented Computing.However, most of research only describes the semantic of Web services from a singleperspective of the interface functions, while ignoring the semantic information ofspatial-temporal, context and performance, especially lack sufficient attention to thenature of service interaction, the relationship between the services and their effect onService Oriented Computing, thus hampering the automatic discovery andcomposition of Web services.In this paper, the multidimensional semantics given social attribute are introducedinto organization and management of Web services, the Semantic Web, SocialComputing and Complex Network theory are incorporated into service computingmodel, in order to build application-centric Web Services Society model as a coreresearch objectives. Meanwhile, this paper also deeply investigates the key issues inthe Web service society modeling. The main research achievements are as follows1) Propose the Web Service Society reference conceptual model and architecture.By considering the Web service semantic information of functions, spatial-temporal,context and interaction relationship, this paper has designed the Web Service Societyreference conceptual model follow the principle of dynamic growth and priority plot.We proposed the new concept of Service Ontology, that is use the ontology to modelthe multidimensional semantic of Web service. With analyzing the componentelements and implementation principle of Web Service Society model, this paper putforward the Web Service Society architecture for guidance the construction ofmultidimensional semantic-aware Web Service Society.2) Design the automatic Web service clustering method based on socialcomputing. By investigating the different collaborative intention betweenco-occurring tags, this paper designed the social tag network based on theco-occurring relationships. The Web services automatic clustering problem istransformed into a community discovery problem based on society tag network.According to the topological structure of this network, this paper introduce a novelWeb service clustering algorithm based on the concept of subgraph intensity and coherence. Finally the experimental result proves that the proposed method wascorrect and effective.3) Research on the automatic generation and evolution mechanism of serviceontology. Integrating the natural language processing and semantic Web technology,an automatic generation and evolution methods of service ontology have beenproposed based on the analysis of the WSDL document structure and itscorresponding relationship with the ontology. The experimental result shows that thismethod has high accuracy and practicability. The construction service ontologyultimately realizes the comprehensive, multi-level and multi-angle semanticdescription for Web services.4) Three different granularities of service relations, including parameter level,operation level and Web service level, are analyzed and defined from the servicecomposition perspective. The isolated Web services are organized into a complexrelation system by the service relations, and realize the semantic awareness betweenWeb services. In addition, this paper has presented the service composition methodbased on service relations. Finally, comparative case studies were provided toillustrate the feasibility and superiority of the proposed approach.In summary, the isolated Web service are organized into multidimensionalsemantic-aware Web Service Society, and proposed the method used to solve the keyissues in the Web Service Society modeling and its application process. It provides anew way and theoretical support for Service-Oriented Computing.