Web Services Composition Based On Ontological Concept Similarity

Along with the age of the Internet coming, people require urgently that the integration of distributed system can be cross-platform and loosely coupled on the network. Web Services have emerged. Web Services's core technologies are XML, SOAP, WSDL, UDDI and it transmit through the Internet standard technology, using standard specificatic XML concept describing. Web Services are loosely coupled software components. Service Description includes all the necessary details interaction with Services, including the message format, transport protocol and location. Web Services are based on Service-Oriented Architecture foundation, and through the interactions between entities such as the service provider, requestor and registration center to implement the invoked service. Web Services are independent, self-contained, self-described, modular software components. It can be published, found, and called. There is strong portability and interoperability.Web Service is convenient to application program integration, but the function of single service is limited, and it can't satisfy the actual requirement. We have to composite the services that existed into new service to supply more and more functions, so that Web Service can actually play potential. With the abroad application of Web Services, Web Service composition technology has triggered a considerable number of research efforts both in academia and in industry. Combining Web Services that had existed into greater granularity web applications is direct goal in Web Service research. Web Service composition has become an important research subject.Web Service composition means that through service search and interface integration composite some autonomic Web Services according to the requirement, in order to supply new and powerful Web Service. In other word, it is a kind of technology that chooses rather simple and useful Web Services in the Internet, then composites them into new service. Nowadays, there are too many disadvantages among the methods of Web Service description.This paper researches the methods based on semantic web, using OWL-S to extend Web Services, simplifying Web Services into a simple data structure, to solove the problems of Web Services composition.This paper chooses OWL-S as a description of Web Services model, mainly because Web Service composition is facing with many problems, such as a simple Web Service lack semantics, how to composite Web Services, whether the service after being composited equivalent to the service expected and so on, but OWL-S has a formal semantics and has an ability to express strong enough. Semantic Web Services composition becomes possible because of the approach of OWL-S. OWL-S provides a core collection of markup language set, so it can let a computer describe the Web Services'property and function as a way of explaining and excuting. In particular,it describes non-functional properties which most methods overlook,such as QoS property, the user's preferences, constraints, security, dependency, and performance.Using the basic model before as the framework, it can add a description of precondition and effect on the Web Services'description and enables the services to be automatically carried out better in matching and composition.First of all, in order to give a measure for the function and the property of Web Services, this paper introduces the concept of the similarity and the association, matching similarity between inputs and outputs of Web Service. Today the calculation method of ontological concept similarity is divided into similarity calculation based on the concept and similarity calculation based on rule. However, the calculation method based on the rules and reasoning relies on artificial participation or simple heuristic rules. It is not realistic in the large composition. Similarity calculation based on the concept's disadvantage is too complexity. At the same time, these two kinds of methods did not consider the concept of property and role of relations to describe the concepts. Therefore, this paper also improves ontological concept similarity calculation method, considering their advantages, combining these two algorithms into an improved algorithm.This paper set different weights for two similarity algorithms. In the calculation, results of two algorithms are combined in proportion to the final result. This paper uses semantic similarity to quantify the matching degree between service's attributes and the concept of parameters, making better composition matching of Web services, consequently let OWL-S describe Web services better.Secondly, this paper extends the definition of Web Services, and defines a new concept of Web Services and other concepts related to Web Services, such as association of Web Service and so on. At the same time, this paper proposes a message matching algorithm and then proposes a Web Service composition algorithm based on ontological concept similarity. The analysis of the algorithm can not only match the various services, but also enrich the semantic information of Web Services. On the basis of it, Web Services compositon is completed, and get a sequence of Web Services compositon to get the purpose of using semantic ontology to research Web Services compositon. Subsequently, this paper uses a complete example of the process of buying books online to compare the Web Services composition algorithm of considering precondition and effect with the Web Services composition algorithm of not considering precondition and effect to find out the difference between two algorithms and prove the algorithm's validity and correctness.Finally, this paper analyzes the compostion process of house rental information service, and implements a Web Services compositon simulator on Eclipse platform. This simulator uses the proposed algorithm to complete the functions of service composition.The technology of Web Services develops rapidly, the new demand for new issues come out one after another. The Semantic Web theory is mature, according to the new problem, we should use the Web Service composition to solove them timely and accurately. The Semantic Web theory also has its weaknesses. Therefore, in the future, we must also pay attention to other technology and other research methods,so we can improve the Semantic Web theory.