Abstract: In addition to the dynamic of service’ generation and died，the quality differences between provided services，there are a mass of malicious，camouflage，false untrue services in the Internet.As result of that，it is difficult for client to get fast and high-quality services.In this paper，a service-composition framework based on the trust evolution and trust alliance has been proposed.It is different from traditional service-composition model.The trust relationship will be built after the trust evolution between service requester and provider.The trust service requester and provider is going to form a alliance through the relationship，so the malicious，false services will be excluded from the system and the service-composition is carried out in a trust system.On this basis，a services selection algorithm based on credit evolution and alliance is put forward，avoiding the deficiencies of multi-dimensional trust indicators simply weighted in the previous studies.Theoretical analysis and experimental comparison confirm that the framework proposed in this paper has a better ability to organize services，so as to make higher-quality，faster combination of services.

Key words: services composition, trust evolution, information entropy, services selection, Quality of Services

摘要： 除了Web服务本身的动态性和提供的服务质量有差异外，互联网上各种恶意、伪装、虚假的不真实Web服务导致用户很难快速获得高质量的服务。提出了一种面向信任演化与联盟的服务组合框架，将服务提供者与服务请求经过演化而形成一种信任关系，通过组合过程形成的信任关系再将可信的服务提供者与服务请求者演化成一种联盟关系，从而将恶意、虚假的服务排除在系统之外，使得服务组合在可信的场景中进行；在此基础上，提出了基于信任演化与联盟的服务组合选取算法，避免了以往研究中对多维可信指标简单加权的不足。经过理论分析与实验对比证实了提出的框架具有很好的服务组织质量和速度。

关键词: 服务组合, 可信演化, 信息熵, 服务选取, 服务质量