摘要
推荐系统是电子商务系统中最重要的技术之一,用户相似性度量方法是影响推荐算法准确率高低的关键因素。针对用户评分数据极端稀疏情况下传统相似性度量方法的不足,提出了一种基于群体兴趣偏好度的协同过滤推荐算法,根据群体兴趣偏好度来预测用户对未评分项目的评分,在此基础上再采用传统的相似性度量方法计算目标用户的最近邻居。实验结果表明,该算法可以有效解决用户评分数据极端稀疏情况下传统相似性度量方法存在的问题,显著提高推荐系统的推荐质量。
Recommendation system is one of the most important technologies applied in e-commerce.Similarity measuring method is fundamental to collaborative filtering algorithm.Traditional similarity measure methods work poor when the user rating data are extremely sparse.To address this issue a novel collaborative filtering algorithm based on group interest preference degree is proposed.This method predicts item ratings that users have not rated by group interest preference degree,then a traditional similarity measure is used to find the target user＇s neighbors.The experiment results suggest that this method can efficiently improve the extreme sparsity of user rating data,provide better recommendation results than traditional collaborative filtering algorithms.
出处
《计算机工程与应用》
CSCD
北大核心
2010年第34期129-131,共3页
Computer Engineering and Applications
基金
国家科技支撑计划资助项目(No.2008BAH21B03)
山东省高等学校科技计划项目(No.J08LB68)
关键词
协同过滤
群体兴趣偏好度
平均绝对偏差
collaborative filtering
group interest preference degree
Mean Absolute Error（MAE）
