基于多级图划分的协同过滤算法研究

摘要

为克服传统协同过滤推荐算法中存在的局限性,提出了基于多级图划分的协同过滤推荐算法. 该算法在对系统中存在的产品使用多级图划分算法进行聚类的基础上应用协同过滤推荐算法对用户进行推荐. 实验结果证明,该算法可以在对推荐准确率影响较小的同时有效提高推荐系统的效率.%To solve the limitations of the collaborative filtering , it proposes collaborative filtering algorithm based on multilevel k-way graph partitioning .This algorithm uses multilevel k -way graph partitioning to cluster i-tems, predictions are then computed independently within each partition .The experiment shows that this algo-rithm can greatly improve the efficiency of system by slightly decrease the accuracy of prediction .