The Research Of Data Mining And XML

Recently, with the development of Manager Information System and E-business, the problem of data-rich but knowledge-poor is increasingly severity .How to find the useful information for enterprise in data have desiderated to be solved. Data Mining have become the focus of research in fields such as AI and Database, etc. In this paper, the detail of Data Mining is introduced comprehensively, and we bring forward the approach to the problem of data integration in Data Mining with the techniques of XML. Moreover, the PMML is deeply analyzed.In the commencement of Data Mining, we will build the Data Warehouse. In the part of Data Warehouse research, the process of building Data Warehouse, the granularity, the management of metadata, are specially introduced in this paper. And then the technique of Data Mining, the language of Data Mining, the representative model and the different arithmetic in models are introduced in detail.In this paper, we adopt the mobile enterprise DSS as the antetype system, and with the techniques of Data Warehouse and Data Mining, design the system structure. And with B/S mode, we have implemented the subsystem of Data Mining.We use the typical algorithm of Association Rule Mining in this paper, Apriori algorithm. Then we analyse algorithms of Association Rule Mining and its merit and shortcoming, and give a method to Association Rule Mining implemented in Data Query Language using the technique of Hash. The performance study shows that the program is efficient.The different structure database is the problem of data integration. We research the difficult problem of data integration in Mining, and put forward to solve the problem with the technology of XML.With the development of XML technique, Predictive Model Markup Language comes forth, it is the description and definition language of data mining model, so the data mining system have the criterion in defining anddescribing model,and all kinds of the data mining system can share model. Using PMML to building model, we can embed the data mining model in other application system. Finally, we parse the PMML 2.0 -- DTD of Association Rules Model.