针对服务质量波动所造成的选择风险性,提出一种支持用户风险偏好的服务评估方法.首先,根据区间数理论处理缺失信息,对用户评分进行完整描述.然后,根据用户面对不确定信息的风险倾向性不同,引入风险偏好对用户进行细分,并确定用户的服务感知风险.最后,根据服务收益最大化原则,将服务属性进行划分并结合用户属性偏好度得到服务收益,权衡服务收益与用户感知风险得到综合评估结果.仿真实验表明:与其他服务评估方法相比,本方法在表达用户属性偏好的同时也清晰表达了用户的风险偏好特征,更好地体现用户个性化差异,具有更高的用户满意度.
We propose a service evaluation method based on risk preferences (SEM-RP) for the risk caused by quality volatility. First, complete description rating information according to the interval number theory is made. Then we use risk preference to identify user' s perceived risk in order to express different user' s risk propensity. After that, service revenue is obtained by calculate service attributes and users' preferences according to the principle of revenue maximization, and finally we get service evaluation rating by calculate service revenue and user' s perceived risk. Simulation experiments and results demonstrate that the proposed method can clearly express the characteristics of user' s risk preference while expressing user preferences attributes and which have the performance of higher user satisfaction.