一种QoS可信增强的服务选择方法  

A QoS Trustworthiness Enhanced Service Selection Method

作　　者：贺兴亚[1] 王海艳[2] 杨文彬[2] 

机构地区：[1]扬州大学信息工程学院,江苏扬州225009 [2]南京邮电大学计算机学院,江苏南京210003

出　　处：《武汉大学学报（理学版）》2013年第5期443-448,共6页Journal of Wuhan University:Natural Science Edition

基　　金：国家自然科学基金资助项目(61201163);江苏省自然科学基金滚动项目(BK2011072);江苏省博士后科研资助计划资助项目(1002005C)

摘　　要：现有的服务选择方法在标准化服务质量(quality of service,QoS)属性时,忽略了QoS的可信性,候选服务仅仅基于QoS值进行排序,很大程度上影响了选择结果的可信性及精确性.为解决以上问题,提出了一种QoS可信增强的服务选择方法,使用概率论的相关知识,将QoS的可信性问题转变为利用样本观测值检验属性值的事后检验问题,在保证QoS属性值可信性的基础上,对其进行标准化处理,进而计算用户对服务各个属性的满意度,选择出最能满足其实际需求的候选服务.应用实例与分析表明了该方法的有效性与可行性.Traditional service selection methods have not paid much attention to the trustworthiness of QoS attributes during the normalization process.Candidate services are ranked simply based on their QoS values,which badly restricts the trustworthiness and accuracy of selection result.To address this problem,a QoS trustworthiness enhanced service selection method is proposed.With the employment of probability theory,trustworthiness of QoS attributes are transformed into the back testing of the sample QoS attributes.Normalization of QoS attributes are realized based on the trustworthiness of these attributes.After calculating the satisfaction degree on each attribute,appropriate service which best meets the requirements are selected.An application example together with the associated analysis demonstrates the feasibility and effectiveness of the proposed method.

关 键 词：服务选择 服务质量 可信性 前景理论 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]