Abstract: By analyzing sparsing and cold-start problems of Item-based collaborative filtering algorithm, an Item-based collaborative filtering algorithm using attribute similarity was proposed. This new algorithm revises the original similarity using item attribute, takes synthetically into account the influence of item attribute and user rating. It improved the measure results in the face of data sparity and new item recommendation.

摘要： 通过分析传统Item-based协同过滤推荐中的稀疏性问题以及新项目的冷开始问题，提出了一个基于属性相似性的Item-based协同过滤算法。该算法利用项目属性的相似性来修正原始相似性计算，综合考虑项目属性和用户评价对推荐的影响，改进了传统相似性度量方法在评价数据稀疏和新项目推荐中测量结果不够准确的问题。