基于工作流的数据挖掘PMML模型实现

随着计算机、数据仓库等技术的发展,数据挖掘在电信、银行等行业中得到了越来越广泛的应用.本文提出了应用PMML标准实现不同数据挖掘模型、工具间数据的传递的方法,同时将工作流应用到数据挖掘系统中,实现了由用户定制数据挖掘过程和选择算法的功能,解决了多个数据挖掘模型、算法间的互操作问题,同时运用工作流增强了数据挖掘系统用户自主定制和选择的能力.

An Implementation of Data Mining PMML Based on Work Flow

Data Mining plays a more and more important role in the research areas in the telecommunication and financial industries as the computer and database technologies are becoming mature.The paper proposes a data transfer method among different data mining tools by applying the PMML standards and integrates job flow (or control flow) into the data mining system to accomplish the specifications of data mining processes and the functions of algorithm selection. The method resolves the co-operation of multiple data mining systems and provides the flexibility for the users to define and select the data mining systems by integrating the control flow.