The Research On Matching Operation In Model Management Technologies

Schema matching is one of core operators in the technology of model management, and has become an important part of model management. The technologies of schema matching in model management are analyzed synthetically, and then two algorithms for schema matching are proposed in this paper and are applied to a prototype which is designed to deal with the update and maintenance after pubulishing XML views, the concrete contents are as follows.Firstly, SMD (Schema Matching using Duplicates) algorithm is analyzed, according to its shortcomings and limitations, ISMD (Integrated Schema Matching using Duplicates) algorithm is proposed. Then ISMD algorithm is analyzed and verifyied respectively through instance and experiment, which can improve correctness effectively.Secondly, the algorithm for schema matching with opaque column names and data values is analyzed. According to the large computation, the nodes filer algorithm is proposed, which can filter the irrational nodes and reduce the computation of mutual information. Based on these, the algorithm for schema matching with opaque column names and data values is improved, and then SMOCD (Schema Matching with Opaque Column Names and Data Values) algorithm is proposed, which deduces the time complexity and can find the matching nodes correctly compared with the primal one.Thirdly, appling the basic idea of model management, a prototype is designed to deal with the update and maintenance after pubulishing XML views to realize the update and maintenance of published XML views automaticly.Finally, the algorithms proposed in the paper are validated throughextensive experiments.