Abstract: Collaborative filtering is the most successful and widely used recommendation technology in E-commerce recommender systems. However, traditional collaborative filtering algorithm faces severe challenge of sparse user ratings and real-time recommendation. Aiming at the problem of data sparsity for collaborative filtering, a high efficient personalization recommendation algorithm based on nearest neighbor is proposed. The algorithm refines the user ratings data using dimensionality reduction, uses a new similarity measure to find the target users’neighbors, and generates recommendations. The experimental results argue that the algorithm efficiently improves sparsity of rating data, and provides better recommendation results than traditional collaborative filtering algorithms.

Key words: recommendation system, nearest neighbor, user similarity, dimensionality reduction

摘要： 协同过滤是目前电子商务推荐系统中广泛应用的最成功的推荐技术，但面临严峻的用户评分数据稀疏性和推荐实时性挑战。针对协同过滤中的数据稀疏问题，提出了一种基于最近邻的个性化推荐算法。通过维数简化技术对评分矩阵进行优化，降低数据稀疏性；采用一种新颖的相似性度量方法计算目标用户的最近邻居，产生推荐预测。实验结果表明，该算法有效地解决了数据稀疏，提高了推荐系统的推荐质量。

关键词: 推荐系统, 最近邻, 用户相似性, 维数简化