融合用户和项目相关信息的协同过滤算法研究

Research on Collaborative Filtering Algorithm Based on Fusing User and Item's Correlative Information

针对User-based协同过滤和Item-based协同过滤算法的不足,提出了一种新的推荐算法。该算法融合用户-项目评分数据集所包含的用户相关和项目相关的信息来推荐商品,并且利用模糊聚类技术分别将相似的项目和相似的用户聚类,改善传统推荐算法的数据稀疏性和可扩展性问题。实验结果表明,将用户相关和项目相关的信息融合能够提供更好的推荐。

Aiming at the disadvantages of user-based collaborative filtering and item-based collaborative filtering algorithms, the paper proposed a novel recommendation algorithm that generated item＇s recommendation by fusing user and item＇s correlative information inhering in the user-item rating dataset. The algorithm also involved the fuzzy clustering of similar items and similar users to improve the data sparsity and scalability of traditional collaborative filtering algorithms. Experiments showed a better recommendation could be provided by fusing user and item＇s correlative information.