云计算环境下可信服务的个性化推荐框架

Personalized Recommendation Framework for Trustworthy Services in Cloud Paradigm

针对当前云计算环境下服务选择中服务推荐技术的不足,提出一种面向可信云服务的个性化推荐框架(PerReF),该框架基于独立、开放的云评价中心和云推荐中心架构,以可信云服务的多属性分析和概率统计分析为基础,使用模糊综合方法集成消费者对云服务可信属性的历史评价,并结合潜在用户在不同应用场景下的个性化需求,从可信属性权重、可信度期望、成本期望角度出发,对云服务进行过滤,再通过多次迭代的模糊相似度计算,找出最适合用户个性化特征的云服务.仿真实验分析表明,PerReF能够适应复杂的云计算环境,提供具有较高用户满意度的个性化推荐结果.

Aiming to the service recommendation problem in cloud paradigm, this paper proposed a personalized recommendation framework for trustworthy services in cloud computing environment-PerReF, which consists of the independent, open cloud evaluation center and cloud recommendation center. In it, based on the multi-attributes analysis of trustworthy service and the probability statis- tics, the service evaluation data of trust attributes submitted by consumers was integrated by fuzzy synthetic method. Considering the personalized demands from potential user in specified scenery, it＇s necessary to filter some cloud services with bad quality of service according to the weights of trust attributes, trust expectation and cost expectation. By means of the results about the fuzzy similarity computing, the most suitable service matching the personalized features was recommended to potential user. The simulation experi- ments showed that the PerReF is suitable to the complex cloud computing environment, and can provide the personalized recommen- dation results to achieve the high user satisfaction.