Research On The Service Model And Methods Based On Ontology

Service computing is a landmark of distributed systems and software integration areas,it is a network component of autonomous,open and it is independent of idiographic language environment,it make the distributed system is reused,flexible and scaleable.Service discovering is important in service discovering,user can not get satisfied service if there haven't service discovering efficiency.UDDI(Universal Description Discovery Integration)is a center of service registration,it provides a mechanism of service publication and discovering,however,a good methods of service discovering should based on semantic,thus,user can get their service satisfied.To get the satisfied service based on semantic becomes the goal of service discovering and selection.Collaboration,negotiation and service scheduling are also playing important roles of service system.Ontology is an important tool of knowledge express and information share,it is development from the traditional knowledge express methods,and it is used widely in knowledge processing area.In order to select and discover service based on semantic,it is significance to use ontology in service discovering and service selection.In this paper, we develop some relevant researches from psychological views,we propose a similarity measurement method based on semantic units(SMBSU),we research a model of service discovering and selection based on ontology(SMBO),and also we research the collaboration and negotiation of the model.Finally we research the service scheduling methods based on semantic and market.There are the main works and innovations(1)We advance the methods of concept similarity measurement based on semantic units (SMBSU),and we get the methods to measure the similarity of concepts,relations and ontologies.Because of the flaw of previous models,we extend the concept attributes by semantic and propose a method to measure the similarity based on the semantic unit.We use the semantic unit to express the concept implication,and we use the support degree to show the contributions of the semantic unit.We do successfully the concept similarity measure by considering the relevance,resemblance,non-symmetry,and the support degree.We look on the relation as concept,and we can measure the ontology similarity based on semantic unit.Finally,the experiment proves that our method is better than other methods.(2)We explore an model of service discovering and getting based on ontology(SMBO)Discussing the methods and faults of the resource organization and service math in grid environment,to share the resource organized by ontology,this paper explore the methods of service discovering and service providing based on semantic relations in semantic grid environment.We propose the bus and star model to discover and gain service.The resources are shared in semantic level by OWL description logic and the constraint satisfied.There provide a flexibility and an intelligent method to integrate the resource and to gain the service.Finally,an example shows the process of the service discovering and getting.(3)We research the methods of collaboration in service discovering and selection,and we discuss the conflicts of service selection in pervasive computing and we use the negotiation to solve the conflicts.Lacking of the connecting between ontologies,single ontology is easy to be "Isolate Island",user can get services satisfied based on those ontologies.Based on SMBSU and SMBO,we research the collaboration method between ontologies.Ontologies can integrate by collaboration,and we can find the hide relations between ontologies,a case shows the collaborative process. In service selections,lots of ontologies may ask for one resource,there may have conflictions in the service selections.By the methods and ideas of negotiations of Agents,we research two methods to solve the problems of service selections in pervasive computing;the first is the Votes methods and the second is the minimum distance methods.The idea of votes methods is to select the resource by the numbers of votes,when the similarity number reach the assumed number, the votes of this resource are add one,we select the resource of which the votes are maximum.The idea of minimum distance is to select the resource of which the distance is shortest.The distance is calculated by the similarity number and the best satisfied number.We also research the principles of offsets value.Our method is that the longer of the distance,the more of the offsets value.(4)We research a methods to schedule the service based on semantic and marketWe have a comparison of some present methods of resource scheduling firstly, and then we discuss the reason and importance of the market mechanism,and also we discuss the methods of finding the best solution of the resource allocation.We design a method to measure the similarity of ontology based on semantic elements. We propose a model of service scheduling On the basis of the market and semantic matching(SSBMO).There provides a method to combine the market and semantic, there have a method of computing of the ontology similarity,the selecting of the efficiency functions,the decision of the price of the resource,and the solution of the resource allocation on the constraint condition.Finally,we have a comparison of the model between Max-Semantic and Semantic-Cost-Max-Min,it shows that our methods can not only allocate the resource,but also it can provide the service user satisfied.