摘要： 当前的推荐方法未能从个性化效用角度评价推荐项目，因此用户需按自己的偏好，在推荐结果中进行再次筛选。针对该情况，提出一种基于效用的个性化推荐方法。该方法采用逼近于理想值的排序法(TOPSIS)作为衡量推荐对象效用的基本方法。为克服TOPSIS中静态权重设置的不足，采用可变精度粗糙集发现用户对属性的偏好。实验结果表明，该方法能为用户提供更好的个性化效用及准确性的推荐服务。

关键词: 个性化推荐, 多属性效用, 变精度粗糙集, 推荐服务

Abstract: Current recommendation methods lack of ability to evaluate utility of the recommendations according to user’s preferences. So users have to make a choice among recommendations. On the basis of this, this paper presents a personalized recommendation method based on utility according to user’s preferences. This method uses TOPSIS to evaluate utility of recommendations. In order to overcome the shortcoming of static weight in TOPSIS, it adopts variable rough set to discover user’s preference for attributes. Experimental results show that the method can provide more appropriate recommendation service with better utility for users.

Key words: personalized recommendation, multi-attribute utility, variable precision rough set, recommendation service

中图分类号: 

• TP311