基于质量相似度的Web服务信誉度量  

Web service reputation measurement based on quality of service similarity

作　　者：付晓东[1] 邹平[2] 姜瑛[1] 

机构地区：[1]昆明理工大学信息工程与自动化学院,云南昆明650051 [2]昆明理工大学管理与经济学院

出　　处：《计算机集成制造系统》2008年第3期619-624,共6页Computer Integrated Manufacturing Systems

基　　金：国家自然科学基金资助项目(60703116);云南省应用基础研究计划资助项目(2006F0024Q);云南省教育厅科学研究基金资助项目(6Y0080D)~~

摘　　要：为提高Web服务信誉度量的灵活性和真实性,提出一种基于服务质量公告值和实际值相似度的Web服务信誉度量模型。通过上、下两层本体,描述了Web服务服务质量的语义和层次化的服务质量指标体系。提出一种支持层次化服务质量指标体系的质量相似度算法,并将质量相似度作为信誉度量的依据。信誉度量算法以统计理论为基础,引入遗忘因子和波动因子,并考虑了服务质量的发展趋势和波动情况对服务信誉的影响。实验表明,该模型具有良好的灵活性和实用性。To improve the flexibility and authenticity of Web service reputation measurement, a Web service reputation measurement model based on the similarity of advertised and factual Quality Of Service （QoS） was proposed. Upper and lower QoS ontology were designed to describe the semantic of QoS and the hierarchical QoS item system. Similarity algorithm for QoS was proposed which supported the hierarchical QoS. Subsequently, a statistics-theorybased Web service reputation measurement method was put forward which using QoS similarity to compute Web service reputation. Evolution trend and fluctuation of QoS were taken into consideration in reputation measurement with forgetting factor and fluctuating factor. Experiments showed that this model was flexible and practical.

关 键 词：WEB服务 服务质量 相似度 信誉 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]