摘要
传统的Web服务主要是基于UDDI的技术规范,只提供了按照简单分类和关键字的服务发现方法,造成查准率与查全率低,影响服务复用和服务组合的相容性。带有语义的Web服务通常也只关注服务功能性的匹配,因缺乏服务质量描述和灵活、有效的服务匹配方法,而难以保证服务的全面匹配和快速定位。本文分析了现有的一些相关研究,在Web服务模型的基础上,结合“逐渐逼近”的思想,提出一个三层次的匹配筛选算法,并对基本描述、IOPE、服务质量各层的匹配算法进行了深入讨论,用相似函数来度量服务提供方与服务请求方的Web服务相似程度,为Web服务发现提供了一种有效的方法。
Traditional Web services only provide simple classification and keywords-based methods based on UDDI technical specification mainly, lead to low precision and recall, compatibility is influenced of service reuse and composition. Web services with semantics also always set focus on function matching, thorough matching and quick location are guaranteed difficulty without quality of service and vivid , efficiency matching method. In this paper, existing related researches are analyzed, on Web service model, combine the thought of＂approach gradually＂, a kind of matching and filter algorithm with three level is presented, each of algorithm, namely basic description, lOPE and QoS matching is thoroughly discussed, similarity functions is introduced to measure Web service similarity degree of provider and requester, this approach will be significant to Web services discovery.
出处
《计算机科学》
CSCD
北大核心
2007年第6期116-121,共6页
Computer Science
基金
国家自然科学基金项目(60573010)
云南楚雄师院基金项目(05-YJQN02)。
关键词
服务发现
服务匹配
服务质量
相似函数
相离度
Services discovery, Services matching, Quality of service, Similarity functions, Deviation degree
