Research On Ontology Based Semantic Web Services Composition Approach

With the rapid growth of the web services'quantity, the problem about how to discover, composite and implement web services effectively in order to improve the reuse rate of web services, reduce the development costs and to improve the competitiveness of enterprises has become an important question in the academia and industry field which needs to be resolved urgently.In the network environment, people's demand for personalized services is growing continuously. Personalized service is a network information service, which is mainly based on the user's settings through the use of computers and network technology to collect, arrange and analyse the personal information resources of user and then to recommend related services to meet the needs of the user's individual requirements. Meeting user's service requests also needs to consider the user's personal preferences during the recommendation of web services or composited web services. Besides, in order to provide proper services to meet the user's individual requirements, it has to improve web services composition efficiency and the success rate of the system. This paper makes a meaningful attempt to solve these problems.At present, the common service description language is WSDL which is lack of semantic information. At the same time, the current UDDI only supports keyword-based service discovery, which obviously can not meet the user's individual requirements. So how to use the user's information to find the web services and then composite them which can meet the personalized needs of users has become the key problem of web services composition. The semantic web adds semantic information to the discription of web services and enables computers to collaborate well with each other and with people. Such semantic web service focuses on intelligent logic reasoning, automatic discovery and automatic compostion of web services.In order to achieve the automatic discovery and automatic composition of web services and to improve the composition efficiency and the success rate to provide the users with personalized services, this paper is mainly based on domain ontology to do the following research:Firstly, the paper introduces the research background and significance of the current composition of web services, and analyzes the problems in this field, and then provides an overview of the main job of this papaer. Secondly, the paper analyzes the current web service standards and the important technologies such as web service protocol SOAP, the web service description language WSDL and UDDI. And then gives the introduction of the Semantic Web, and gives the hierarchy of the semantic web. Then the paper illustrates the concept of ontology, the ontology discription language OWL and the semantic web services description language OWL-S. At last the paper introduces the basic concept of discovery and composition of web services.Thirdly, the paper researches and gives the heuristic composition method which is based on the similarity degree of ontology concept and the user's preferences. The premise of this paper is that domain experts have described a domain ontology using OWL language, and built a specific field's atomic services data base by the use of OWL-S. A calculation method of the similarity degree of ontology concept is introduced, and then by this way the paper calculates the match degree between user service requests and the atomic services or service templates or the precursor services. This paper trys to record user's ontology concepts preference information. The methods of record have two ways: Firstly, when users register in the system , they evaluate the concept preference of domain ontology initiatively; Secondly, when users use the systems, the system will record the input and output information of users'service requests, and update the user's preference table of ontology concept dynamically. This paper regards the process of web services composition as a heuristic map search problem, in order to enable web services composition to reflect the user's preference information, and to improve customers'satisfaction this paper designed a heuristic function. Then the heuristic composition method which is baesd on the semantic similarity degree of ontology concept and user preferences is given. This composition method uses the degree of semantic similarity to search precursor services,and then uses heuristic function to select a proper precursor service so that we can carry out heuristic graph search. Because the result of the web serices composition may be used again, the paper uses service template database technology to save the composition results as a template into each user's individual database so that the user can find the template directly, so this method can improve the efficiency of the web service composition system.The personalized database includes user personalized template database and user's preference table of ontology concept,which can update dynamically in the process of users'access to the system. In order to ensure customers'satisfaction and system efficiency, in this paper, the author designes a personalized database optimizer which can optimize users'template database.Fourth, this paper designs and realizes the web services composition system. The author divides the composition system into three layers: the first layer is the user access layer, the second layer is services composition logic layer and the third layer is the data layer. At the same time, the class diagrams of every layer of are also given. Through analyzing the result of the simulation test the paper proves the validation and applicability of this method in improving the success rate, system efficiency and customer satisfaction of the services composition. During the test, the author researches the OWL-S API of MD University, and calls a practical semantic web service so as to demonstrate the implementation of a semantic web service.