Abstract Customer satisfaction is one of the important indicators to measure web service composition. To overcome the shortage of customer satisfaction in web services composition, the article proposes a personalized web service composition algorithm based on CBR and multi-agents. Founded on the definition of web services customer satisfaction attributes, an attribute preference weight vector solving model is built. Moreover, by searching for service combination optimal solution through multi-objective planning, the personalization of web service combination is realized. A prototype system has demonstrated its feasibility.%客户的满意度是衡量Web服务组合的重要指标之一.为解决Web服务组合中客户满意度的缺失问题,提出一种基于CBR( Case-based Reasoning)和多Agent的个性化的Web服务组合算法.在给出Web服务客户满意度属性的定义基础之上,建立了属性偏好的权重向量求解模型,并通过多目标规划寻找服务组合的最优解,从而实现了Web服务组合的个性化.原型系统表明了其可行性.
展开▼