随着网络和智能移动终端的快速发展，为了适应大数据环境，云计算已经作为一种软件范型，根据需求动态的为用户提供服务，而SaaS服务之间协同计算与透明交互过程使得用户的隐私数据极易泄露。本项目面向SaaS服务之间的交互特征，提出了一种云服务协同交互的隐私暴露检测方法。首先，对云服务组合流程BPEL及其执行引擎进行扩展，使其满足运行时隐私暴露监督需求；其次，生成隐私协议可执行片段，插入至相应的扩展活动中，并对插入后的BPEL的正确性和一致性进行验证；再次，通过运行时监督，记录违反隐私协议的云服务参与者，同时对不满足用户隐私需求的云服务进行替换，并利用事务机制对BPEL的演化进行管理。最后，提出了针对服务透明交互与演化的隐私暴露检测算法，开发原型系统，并通过实验验证此方法的可行性和有效性.

With the fast development of internet and intelligent mobile terminal, cloud computing, as a new software paradigm, provides services according to user requirement dynamically in bigdata environment. However, collaboration computing and transparent interaction among SaaS services tends to disclose the user privacy data. In this project, considering interaction characters among SaaS services, we propose privacy disclosure checking method. Firstly, we extend cloud services Business Process Execution Language and its execution engine, to satisfy privacy disclosure supervising requirement. Secondly, we transform privacy agreement to executable program segments, apply program instrumentation to the extended activities, and verify its correctness and consistency. Thirdly, cloud services participants who violate privacy agreement are recorded and substituted through running time monitoring, and BPEL evolution is managed through transaction mechanism. At last, we propose privacy disclosure checking algorithms, develop prototype system, and verify the feasibility and validity of our method through experiment.