文档库

最新最全的文档下载
当前位置:文档库 > 多维度自适应的协同过滤推荐算法

多维度自适应的协同过滤推荐算法

多维度自适应的协同过滤推荐算法

邢哲;梁竞帆;朱青

【期刊名称】《小型微型计算机系统》

【年(卷),期】2011(032)011

【摘要】传统的协同过滤推荐算法明显存在的缺点是数据稀疏性导致所求相似性的不准确,影响最终推荐质量.本文围绕其局限性展开研究,提出一种多维度自适应的协同过滤推荐算法,有机结合三种推荐模型——基于用户、基于项目以及基于评论的相似性计算,将观点挖掘技术运用到协同过滤推荐算法中,并通过动态度量方法自动确定三个维度的权重产生最终推荐.实验结果表明,该算法可以有效缓解用户评分数据稀疏带来的不良影响,提高预测准确率和推荐质量.%Collaborative filtering (CF) is one of the most important algorithms applied in e-commerce recommendation systems. The traditional methods are inefficient when the user rating data is extremely sparse. In order to overcome the limitations, a novel algorithm named MACF (Multi-dimensional Adaptive Collaborative Filtering Recommendation Algorithm) is proposed in this paper. The MACF algorithm creatively combines three recommendation models: user-based CF, item-based CF and review-based CF. It successfully integrates opinion mining technology with collaborative filtering algorithm. In addition, a dynamic measurement approach would help determine the weight of three dimensions: user, item and review, and hence get the final prediction result. The experimental results show that MACF can effectively alleviate