Font Size: a A A

Research On Composite Web Service Matching Method

Posted on:2015-04-29Degree:MasterType:Thesis
Country:ChinaCandidate:H HongFull Text:PDF
GTID:2298330422972081Subject:Computer software and theory
Abstract/Summary:PDF Full Text Request
With the deepening of Web Service applications and the increasing number of WebServices on the network, more and more services with the same or similar functions aresupplied. How to accurately and efficiently find and locate the services which servicerequesters need most from a huge number of Web Services group has become a keyproblem of restricting Web Service applications’ development.Present Web Service discovery methods are mainly based on the syntax andsemantic level. Syntax level matching method exists the problems of structuralheterogeneity and semantic heterogeneity, lacks accurate description of services clientsare interested in and does not support the refinement, generalization and the same levelexpansion semantic query, which leads to low recall and precision. Semantic levelmatching method just describes the static Web Service without covering any relevantservice execution process information and considering the Web Services’ non-functionalattributes. It causes a low precision and the service discovered not to satisfy the userrequirement as well as they want. Because both of the above methods don’t take theexecution process information and QoS requirement information between the subservices of composite Web Services into account, it leads not to accurately judgecomposite Web Services’s similarity level, lowers the service’s precision and doesn’tprovide an effective means to match for composite Web Services. Therefore, researchon composite Web Service matching method in this article has important academicsignificance and practical value.Aiming at the above problems, on the basis of the expansion of OWL-S, this paperadopts the cycle progressive and stepwise refinement ontology modeling method todesign process ontology and gives its formalized definition which enhances QoSsemantic description’s capabilities. Based on the above process ontology this paperabstracts process model with dynamic interaction and process control functions. At last,the composite Web Service matching method is put forward. The basic idea of thismethod is: Firstly, it uses UML model to conduct the composite Web Service processmodeling. On the one hand, it is for describing the integration service. On the otherhand, it is for constructing the correct excuting model of relevant composite servicenamely platform-independent model. Secondly, according to the mapping relationbetween UML activity diagram and process model, it designs the conversion algorithm to convert the UML model into process model based process ontology built by thisarticle. Thirdly, by giving the transformation rules from process model to process graph,the transformation algorithm is designed to transform the process model into processgraph. Finally, by defining the similarity between the process graph, it designs aefficient process graph matching algorithm to realize the process graphs similaritymatching. Through calculating the similarity value of process graphs, this algorithm canjudge the matching degree of composite Web Services corresponding with the processgraphs.According to the above theoretical foundation, this article designs a compositeWeb Service matching framwork. The framwork mainly includes three modules, namelymodel convertion, process graph conversion and process graph matcher. These modulesare coding to achieve by the Java programming language in Eclipse developmentenvironment. Finally, experimental comparision and analysis result verifies that thisarticle is feasible and effective.
Keywords/Search Tags:Web Service, Process Ontology, Unified Modeling Language, ProcessModel, Service Matching
PDF Full Text Request
Related items