The Research On Model Management Operation

The future of the Semantic Web depends on the whether or not we succeed to integrate reliably thousands of online applications, service and database. The applications that tie together these systems are the importance of the research. Such applications deal with the tasks that arise in the context of database design, database integration, database translation, model-driven website management, data warehousing, etc. The purpose of model management research is to reduce the amount of the programming of the metadata-intensive applications. The importance of model management research is the operators using implement some operations. The paper researches and improves two main operators Match and Merge.Firstly, we point out the disadvantages of the SF (Similarity Flooding) schema matching algorithm, and introduce a new schema matching algorithm based on the SF algorithm. The function MatchSearch(S, T) is introduced in the new algorithm to discovery the complex matching that exists between the models. We compute the candidate matching using the Lexical Analysis to deal with simple matching and the complex matching at the same time that may improve the efficiency more or less.Secondly, we describe the algorithm of implementing the Merge operator using the natural language and introduce the conflicts that happen in the averaging process. We introduce a method to deal with the meta-model conflicts through translating the schemas between XML and others.Finally, we describe the merging process of the multiple sources interface schemas, and introduce two merging algorithm used to deal with the regular interface schemas and the irregular interface schemas respectively and prove the correctness of the algorithm using the experiment.