基于QoS综合评价的服务选择方法

A Service Selection Method with QoS Synthetic Evaluation

从海量服务中准确选择出符合用户真实需求的服务变得日益重要,而服务质量的评定在选择服务的过程中至关重要。基于QoS的已有服务选择方法,通常假定服务提供者和使用者给出的QoS数据都是真实可信的,但这一假设实际上很难得到保证。为此,提出了基于QoS综合评价的服务选择方法。该方法针对来自服务提供者的质量属性,基于历史数据的统计对当前QoS数据进行修正;而针对来自服务使用者的质量属性,基于具有时间系数的历史评价计算推荐用户和目标用户之间的相似程度,以权衡其历史评价的可靠程度,进而得出针对该目标用户的最终QoS参考评价。同时,采用熵值法计算用户的主客观合成权重,并基于提供者修正QoS发布信息和用户可靠的QoS参考评价信息完成服务选择。实例验证结果表明,所提出的方法有效可行。

It is increasingly important for users to exactly select a service meeting their requirements from the mass services. Assessment on the quality of service is crucial in the process of service selection. Existed QoS-based service selection methods generally assume that QoS data from service providers and users are trustable. However,it is not realistic. Therefore, a service selection method based on com- prehensive evaluation of QoS is proposed, by which the values of quality attributes provided have been adjusted based on historical statis- tical data and that from user reviews have been integrated into a comprehensive QoS evaluation for a target user based on the user similari- ty computed through reviews on time factors. Moreover, entropy method is adopted to calculate the user＇ s subjective and objective syn- thesis weights with further operation of service selection based on the adjusted publishing QoS from providers and the reliable reviewing QoS from users. The experimental results show that it is effective and feasible.