一种基于用户偏好序列的协同过滤推荐

协同过滤技术被成功地应用于个性化推荐系统中,但随着系统规模的扩大,它不能真实地反映用户的兴趣偏好.针对此缺点,提出了一种新的协同过滤推荐算法,该算法根据用户偏好序列的相似性来搜索目标用户的最近邻居和产生推荐,从而有效地解决了传统协同过滤推荐中过分依赖不能真实反映用户兴趣偏好的用户等级评价的问题,改进了传统协同过滤算法中计算邻居用户的方法.实验结果表明,该算法在个性化推荐系统应用中取得了较好的推荐效果和推荐质量.

A Collaborative Filtering Recommendation Algorithm Based on User's Preference Order

Collaborative filtering is the most successful technology for building recommendation systems.Unfortunately,this method does not reflect user′s interests with the number of users and items.So this paper describes a new algorithm for collaborative filtering;the nearest neighbors of target user can be found based on the similarity of the user′s preference order and produce recommendation.It can be used to solve the problem on severe dependence of traditional collaborative filtering on user s rank rating,which does not reflect user＇s interests. This algorithm may effectively improve the traditional collaborative filtering algorithms used to find the target user＇s neighbors. Experiment results show that the new algorithm performs well in personalized recommendation system.