分布式并行数据挖掘系统的研究与实现

A Data Mining Method based on VisiBroker and Mutil-thread

分布式多层体系结构,一方面可以适用于分布式应用的需要,另一方面可以减少系统的升级、维护费用,提高系统的鲁棒性。并行计算可以提高系统的执行效率。因此,研究分布式数据挖掘和并行数据挖掘问题很有必要性。文章研究了基于多线程的并行数据挖掘,和基于VisiBroker的分布式数据挖掘的实现原理,以及运行实例。

Using the multi-thread model,Authors put and implement the parallel data mining algorithm.In the form of DLL,the algorithm can be used at different development platforms.In order to make the KDD system have more ex-tendibility and scalability,it uses the CORBA model.Through the mid-ware VisiBroker and OSAgent ,This paper puts and implement s the data mining application server KDD_SERVER and it's application.