Research On DataMining Architecture And MetaData Schema Based On Grid

Due to the increased computerization of many industrial, scientific, and public sectors, the amount of available digital electronic data is growing at an unprecedented rate. The effective and efficient management and use of these data, and in particular their transformation into information and knowledge, is considered a key requirement for success in such knowledge-driven sectors. Data mining is the de-facto technology addressing this information need. But the traditional Data mining systems which run on single machine or local cluster can not process the distributed data effectively, efficiently. So here give a solution base on Grid computing.The main works of this paper include two parts. Firstly, Survey of the theory and technologies grid computing. After comprising the architectures of grid nowadays, the deficiencies of these grid architectures are presented. And indicates that to solve the problems of distributed data mining realize the Data mining grid is the right way. By analyzing the requirements of the data mining grid and comparing the architectures of the grids, this paper proposed architecture of the data mining grid. Based on the Globus Toolkit and other open technology and standards, the architecture provides tools and services facilitating the grid-enabling of data mining applications without any intervention on the application side. Critical features of the architecture include flexibility, extensibility, scalability, efficiency, conceptual simplicity and ease of use. Secondly, this paper investigate meta-data schema used for Data mining grid. After the actual data mining program (i.e. a batch-style executable) is uploaded on a grid server and an XML document (i.e. an instance of the meta-data schema) that describes the program is prepared and registered with the underlying grid information services. Users can discovery and execution of the program in the grid environment easily.