Research And Implementation On Key Technology Of Cloud Service Discovery Based On Ontology

The Cloud Computing has gained great concern since it was proposed in2007, and it hasbeen widely applied in various fields. In the cloud environment, all resources are provided inthe form of service to the users. Therefore, the discovery of cloud service has become one ofthe key research topics. At present, few cloud services is provided by several top companies.Users can find the suitable cloud service manually. As the development of network technology,the amount of cloud service will appear in a rising trend. The manual searching method isapparently not an appropriate way. This thesis focuses on the research work for cloud servicediscovery on the aspects of domain ontology and matching method.The main works in this thesis can be summarized as follow:(1)The construction of domain ontologyWe construct the domain ontology of cloud service. We use the Racer to reason thesimilarity of concept in cloud service description and user requirement. By the reasoningresult, we obtain the concept similarity.(2)The design of cloud service discovery methodIn this thesis, we improve the matching algorithm based on ontology. The algorithm canfind missing Cloud Service and filter the irrelevant Cloud Service.We design the different cloud service discovery methods that consult the domainontology. Through analyzing the feature of cloud service, we propose the concept equivalencematching method. Based on the reasoning technology, we propose the concept containingmatching method. We design the concept similarity matching method. We design the conceptsame matching method in the condition of concept same. The designed algorithm caneffectively increase the success rate of service matching.(3)Sort matching resultThe cloud service relevant to user requirement may display the last of the service list. Inorder to solve the problem, we propose two methods of sorting in this research. First, we putforward the sorting method according to the service utility, which means sorting the servicebased on matching level. Second, we propose the sorting method according to the customers’ preference, which means sorting it based on customers’ most concern property.(4)Finally, a Cloud Service Discovery System is finally realized is this thesis.