Development Andapplication On System Software For Intelligent Data Mining

Data mining is a hot topic on contemporary information technology, which has beenwidely used around many fields, for its availability of mining rules, and then accomplishmentin task of prediction. Now there are lots of data mining software, nonetheless, theconfidentiality of processes on product development and technology is strong, in the light ofloading conveniently and application freely by own data mining algorithm, so it is verynecessary that kinds of own data mining software are developed. Therefore, a set of datamining system software is developed based on Visual C++and intelligent data miningtechnology, which could be in the service of practical application for logging mining.Thecrucial content are as follows:1. Solution for system is designed. The software system is designed into three parts, i.e.,data layer, logic layer, while the presentation layer is the part of user performance, accordingto data mining technology, development environment and mainstream support for the system.This scheme follows the typic design of threee parts, which possesses compactness. Itincorporates data layer and logic layer into client of Client/Server(C/S) mode, what’s more,combining the mixed programming skill with engine server of Matlab, which affords thecomputation partly for the processing of intelligent algorithm in logic layer.2. Intelligent Data Mining is developed. System possesses friendly interface, high abilityof visualization, strongly expansibility. It especially develops visualizational modules of user,data management and image management. It uses the mainstream support of data interface,i.e., technology of ActiveX Data Object in the interaction of data layer and logic layer in orderto complete the task of interaction of database SQL Server2005. As to the procedure ofMatlab, it can be embedded in the system. In the aspect of presentation layer, visualizationand interactive function can be realized by plug-in, i.e., ComboBox, ListView, Pie chart, Curve chart and Bar Chart.etc.3. Data mining system matchs the practical application for well logging. Systemcompletes data mining task of a certain interval in a certain well combined with well loggingdata. It shows good logging interpretation results, after steps of data preprocessing, attributereduction, modeling with SVM method and optimization. System possesses thevisualizational function of dynamic logging curve, access and conversion of logging picture,furthermore, the result of data mining effort can be shown by graph.