Research On Metadata Management Tool Based On Semantic Analysis

During the construction of data center, there are no consistent and uniform definition to the mass data from different business systems. We need the metadata management tool to provide a unified view of metadata, and analyse these data sources, operation details and changes of tracking. Traditional metadata management tools are inadequate, lacking of universality and standardization, and having congenital defects in semantic analysis. So the research of metadata management tool based on semantic analysis has important significance.Based on the CWM, we design the metadata management tool. In this paper, the metadata management process is analyzed and designed; We analyze the functional metamodel management requirements, metadata management requirements and semantic analysis requirements; We design the system structure, containing the metamodel management module, metadata management module and semantic analysis module; Using the results of OWL and semantic reasoning, we propose and prove the rules how the metadata converges to OWL, design the semantic inference rule format, and design the semantic inference engine. We introduce the graphical user interface to design the metamodel, the ways to import and export batch metamodel, and the ways to update metamodel incrementally. We also introduce the algorithm of building metadata and relationships. Meanwhile, we design the semantic reasoning engine based on Jena, which supports the custom semantic inference rules and checks the semantic consistency of metamodel. At last, we presents the metadata lineage analysis algorithm, impact analysis algorithm and different analysis algorithm.