基于DMX的数据挖掘算法包原型的设计与实现

Design and Implementation of Data Mining Algorithm Package Prototype Based on DMX

数据挖掘扩展插件DMX(Data Mining eXtension)语言是由微软提出的在SSAS(SQLServer Analysis Services)中实现的符合OLE DB for Data Mining规范的数据挖掘语言,它提供了创建、访问和管理数据挖掘算法和信息的数据存储方式和开源的可扩展的模型调用体系。为了满足目前国内中小企业迫切需求的数据挖掘工具的应用需求,通过分层设计、逐次封装,设计并实现了一种基于DMX语言的特性和功能并具有平台无关性和应用推广性的数据挖掘算法包原型系统。通过公共测试数据验证了该原型系统的可行性和可用性。

Data Mining Extensions is a Microsoft＇ s data mining language which is based on OLE DB for data mining specification and implemented in the SSAS（ SQL Server Analysis Services）. This language provides data storage method and open source extendible model call system for creating, accessing and managing data mining algorithm and information. In order to meet the needs of data mining tools for present domestic SMEs and through hierarchical designing and gradual encapsulating, designs and implements one data mining algorithms package prototype which is based on the feature and function of DMX language and platform independent and application propagable. The paper has verified feasibility and usability of the prototype.