Research On Semantic Web Service And Its Composition Method

As a novel distributed computing model on Web, Web Service enables efficient integration of various data and software on Web by providing an effective mechanism to facilitate interaction across heterogeneous platform when sharing and integrating multifarious data and software application. Built by combination of Web Service and Semantic Web, Semantic Web Service aims at dealing with misunderstanding issue of XML-based web service interoperation derived from the lack of semantic description ability. By sharing and reusing semantic information about web services, semantic web services make exchanged messages between web services understood by computers, and therefore promote the automation of web service discovery, selection, invocation and composition.Web service composition refers to integration of several web services in a harmony way to offer a set of complex application function which could not be supported by a single web service. Discovery, selection, composition and harmonizing the implement of web services, are all transparent to the users. Although current research work has provided some standards and specifications, such as Web Service Description Language (WSDL), Simple Object Access Protocol (SOAP), Universal Description Discovery and Integration (UDDI) etc, some actual problems still hinder the description, discovery, selection and composition of semantic web services in such an open and distributed network environment as followings:(1) How to find the semantic web service which is exactly the client needs. This problem consists of two parts. The first one is how service providers expound the problem the web service can solve on the premise of hiding the underlying technological details; the other one is how the client, who only knows what he/she needs and has no idea where and in which way he/she can find useful information about required web services, can discover web services to satisfy his/her requirement. To the web service requirement described previously, an effective approach is needed to be utilized to realize semantic description of web services.(2) How to effectively and efficiently match between semantic web service description and client requirement:currently web service description explains the interface by depicting the exchanged messages during interaction between the client and the web service, while clients state the requirement by declaring what application function they want. Semantic web service could decrease misunderstanding between service providers and clients and enable information about web service understood by computers by adding some semantic annotations, however these semantic information would largely reduce the efficiency of service matching especially when there are lots of web service candidates existing on web.(3) How to implement flexible and efficient semantic web service composition: sometimes there are no single web service available to meet the client’s complicated application requirement, and at this time composition of several web services which generates one new web service with more value is a feasible solution. However, the difficulties of semantic web service composition derives from open, volatile and unpredictable network environment, collaboration among every component web service, translation from workflow definition to web service composition process description, and so on.Based on the analysis of the practical problems of web service composition, this thesis discusses the limitation of current method and language of web service description, semantic annotation, composition approach. At the same time, we propose Semantic Web Service Community (SWSC) model, and explain the conceptual model, construction structure, service semantic description approach and activities included in SWSC. Additionally, thehesis focuses on semantic web service composition process by improving and optimizing web service discovery and selection algorithms. The main research work presented by the thesis includes as follows:(1)The content, approach and language of web service description are discussed in detail, and meanwhile the conclusion that semantic web service is an inevitable next step of current web service development, has been reached. Subsequently, we explain how to describe web service in semantic level and analysis what the functional and nonfunctional description of web service means to the web service discovery and selection, and find that effectively clustering and separate web service individuals according to the commonness and peculiarity of web services can efficiently improve the efficiency of web service discovery and selection process.(2)The Semantic Web Service Community (SWSC) model is proposed and the description approach of semantic web service in the circumstance of SWSC is presented in the thesis. At the same time, the development and implementation of SWSC model is depicted by explaining the conceptual model, construction, main activities included in SWSC, implementation rules and the prominent characteristics of SWSC model compared with other semantic web service models. (3)By adding some semantic constraints, the current web service function-oriented matching algorithm based on OWL-S is improved to avoid a type of semantic deviation existing in the process of web service discovery and selection. Based on the improvement of this algorithm, the policy and approach of web service discovery and selection in the circumstance of SWSC is presented, and meanwhile how the web service discovery and selection approach in SWSC model improves the efficiency of web service discovery and selection process by shrinking the search space, is depicted combined with the description of SWSC construction.(4)On the basis of analysis of current problems and approaches of web service composition approaches, a dynamic web service composition method in the circumstance of SWSC is proposed. At the same time, a coordination mechanism is introduced in order to reach and maintain relative balance between actual client requirement and available web service resource on the premise that most of clients’requirements are largely satisfied.The main contributions of the thesis include as follows:(1)The Semantic Web Service Community (SWSC) model is proposed and the conceptual model, construction structure, semantic description method of SWSC model is explained.(2)The current web service function matching algorithm based on OWL-S is improved and optimized by adding some semantic restrictions.(3)A web service composition mechanism and approach in the circumstance of SWSC is presented.(4)A coordination mechanism aimed to reach and maintain relative balance between actual client requirement and available web service resource is proposed. The premise of this mechanism is that most of clients’requirements should be largely satisfied.