Research On Semantic Web Service Discovery Algorithm Based-on OWL-S

Web Service is a kind of new self included and described distributed computing model runningon the Web. It can be described, published, found and called on the network, and is a keysupporting technology to realize of SOA. Due to the incomplete and the inconsistency about webservices description, it is very difficult to the machine accurately search services meet the needsof users from massive web services groups, so web services discovery is one of the key problemsrestricting the applying and developing of web services. With the semantic web is proposed, thecombination with the web service can more effectively support automatic discovery, executing,interpreting and combining by the way of rational use of semantic information in web service, inorder to realize of interoperating and seamless integrating between heterogeneous systems.In the aspect of web service description, the WSDL standard is service description based on thegrammatical level, so is lack of powerful semantic description ability. In November2004, OWL-Sbased on DAML is a current standard description on the semantic web service. It describedfunctions, process and access method about services through a set of sub ontology predefinedaccording to the web service vocabulary. The vocabulary defined the complex relationshipbetween the various objects. So the machine can understand the web service and support theQoS.According to the OWL-S Standard Specification, web service ontology is represented by3subontology. ServiceProfile, ServiceModel and ServiceGrounding respectively describe functions,process and access method. The information described by the sub ontology ServiceProfile is themain basis for semantic service discovery, including the basic summary info, IOPE description andnon functional QoS. Based on the OWL-S semantic description, there exist some algorithms, suchas the grades matching algorithms, the similarity algorithms based on semantic distance, thesimilarity algorithms based on concept attributes and the semantic similarity algorithms based onlogic reasoning. The discovery algorithms are mostly on the IOPE matching. However, fewcomprehensive algorithms fully support for the OWL-S ontology service description model.This paper firstly introduces the basic technology about web service discovery, including theweb service architecture, semantic web concepts and the OWL-S specification. These providetechnical basis and support. Then, it mainly analyzes various service matching algorithms aboutsemantic service discovery, including the classic DAML/UDDI grades matching and the semanticsimilarity algorithms based on formal numerical calculating. In the aspect of profile infodescription of the OWL-S service ontology, it analyzes the best matching methods from each layeraccording to the respective characteristics and different effects on the quality of matching of3aspects of the OWL-S in services discovery, in order to we can more accurately search services. According to the traditional UDDI service discovery framework, this paper presents anextended service discovery model based on OWL-S semantic techs. It presents the implementingmechanism of the model in the aspects of process of service registration and request. At thesame time, the comprehensive semantic matching algorithm of the model is presented. Then thesemantic discovery system is realized of through existing tools, in order to prove the valid of theextended model and the semantic matching algorithm. In the end, this paper summarizesshortcomings and deficiencies of the extended model and algorithm, and briefly presents thebasic methods and ideas to solve the problems.