基于知识网格服务的分布式数据挖掘研究

Distributed data mining based on knowledge grid services

广域网中存在地理上分布的海量的各种数据,分析和处理这些数据需要利用高性能的分布式并行处理系统,网格能够满足这种要求。知识网格就是使用基本的网格服务(通信服务、信息服务、授权服务和资源管理服务)去建立特定的分布式并行知识发现工具和服务。结合知识网格特点,讨论了知识网格的体系结构和支持知识挖掘应用的服务集。运用分布式数据挖掘的元学习模型,给出了利用知识网格提供的知识挖掘服务实现分布式数据挖掘的过程。

There are large data sets maintained over geographically distributed sites in WAN. It is necessary to analyze and manage the data by using high-powered distributed and parallel processing system. The grid can meet the need. Knowledge grid uses basic grid ser- vices such as communication, authentication, information and resource management service to build specific distributed and parallel knowledge discovery tools and services. The characteristic of the knowledge grid is summarized, the architecture of the knowledge and service sets supporting knowledge mining application are discussed. Based on meta-learning model of the distributed data mining, the process of realizing distributed data mining using knowledge mining services provided by knowledge grid is offered.