Research And Application On Semantic Web Service Matching Mechanism Based On Load Balancing In SOA

The subject of this paper comes from the Science Research Project"Key Technology Research and Application of the Survey and Design Enterprise Infomationization", which is one of the National Science & Technology Pillar Program during the Eleventh Five-Year Plan Period.Service-oriented architecture (SOA) has changed the software development and application model of the past,it can significantly improve the responsiveness of business operations and enable enterprises get real-time support to the business changes. However, there will be mass of web services with the process of the large-scale implementation of SOA, a complex situation may also arising in the functions and interfaces of the web services released.Therefore, for the consumption requests to web services, how to accurately find the list of services that the users really need in the large numbers of the web services is a serious problem to be solved. The traditional service discovery mechanisms are mostly based on exact query using keywords, so these mechanisms lack of machine-understandable semantic information, the available list of services that the registry center returns to the client according to service discovery rquests form the uers is not accurate and full, so design a effective semantic web service matching mechanism is the key factor to solve these problems.Through the research of traditional semantic similarity algorithm, this paper has accomplished the following tasks:①An improved single-concept semantic similarity algorithm was presented, which can improve the existent shortage of the traditional concept algorithm method based on description logical inference, it can quantifiably calculate the semantic web service's match level to the user's requests, and used the single concept semantic similarity algorithm to further designed the concept collection semantic similarity algorithm.②In order to reduce the additional computing resource expense caused by semantic computing, a improved service match request routing algorithm which based on load balancing in the service registry center was designed, the algorithm can fully use the registry center's computing resource, and improved the throughput rate of the service registry center to execute the service matching tasks.③Based on the research of web service's semantic description framework on this basis, this paper designed the way on how to make semantic description to the web service, which worked from three ways of input, output and classification.④Researched of how to make semantic extension to the traditional web service registration system based on the web service's semantic description framework.⑤Made further efforts to design the framework of semantic web service registry center's entire running environment, and also designed semantic web service's publishing strategy and matching strategy.Based on the experimental data analysis, it shows that the improved semantic similarity method can more effective to accurately compute the concept matching degree, the load balancing system which services the upper registry center can work better to route the matching requests, the utilization of service registry center's overall computing resource is improved, finally, combined with the project, this paper analysis the application of semantic web service matching based on load balancing in the survey and design enterprises.