摘要
为对多个服务注册中心提供支持,方便服务访问日志的记录与挖掘,提出一种基于P2P的分布式服务执行挖掘框架。针对跨组织业务关联的需求,利用该框架构建服务注册联盟机制,设计基于日志库的Web服务关联规则挖掘算法进行组合服务频繁序列挖掘。仿真结果表明,该算法能有效挖掘日志库中的执行与交互信息,提高服务选择与组合效率。
In order to provide support for multiple service registry center,and convenient for recording and mining of service access log,this paper proposes a distributed service executive mining framework based on P2P.Aiming at the requirement of cross-organizational associated business,it constructs service registration union mechanism using this framework.In registration union,it designs Web service associated rule mining algorithm based on log base to progress frequency sequence mining of composite service.Simulation results show that this algorithm can mine executive and interactive information effectively,enhance efficiency of service selection and service composition.
出处
《计算机工程》
CAS
CSCD
北大核心
2010年第9期23-25,共3页
Computer Engineering
基金
国家自然科学基金资助项目(60472093)
国家教育部博士点基金资助项目(200801120007)
关键词
组合服务
关联规则
序列模式
数据挖掘
composite service
association rule
sequence pattern
data mining
