基于RESTFUL的面向服务数据挖掘原型系统的设计与实现

DESIGN AND IMPLEMENTATION OF RESTFUL-BASED SERVICE-ORIENTED DATA MINING PROTOTYPE

目前大多数服务是通过用RPC风格构建的,然而RPC风格的架构在可扩展性、耦合性和安全性等方面存在着瓶颈。结合分布式数据挖掘系统的特点,提出基于RESTFUL服务的分布式数据挖掘架构的思想,依据设计并实现一个对多节点上数据库频繁项集挖掘系统。实例论证了面向服务数据挖掘的合理性及利用RESTFUL服务实现分布式数据挖掘的优势。

At present most services are established with RPC style.However there are bottlenecks at scalability,coupling,security and other fields with RPC style.By integrating the characteristics of distributed data mining system,the paper proposes a distributed data mining architecture based on RESTFUL service.According to the design,it also implements a frequent itemset mining system from multi-node database.Instances demonstrate the reason for service-oriented data mining and the advantages of RESTFUL service to realize distributed data mining.