YANG Haiqing, WU Hao, ZENG Junfei. Algorithm of collaborative filtering recommendation based on non-negative matrix factorization and item heat[J]. 2021, (2): 117-119.
YANG Haiqing, WU Hao, ZENG Junfei. Algorithm of collaborative filtering recommendation based on non-negative matrix factorization and item heat[J]. 2021, (2): 117-119. DOI: 10.13873/J.1000-9787(2021)02-0117-03.
基于非负矩阵分解与项目热度的协同过滤推荐算法
摘要
针对基于邻域的协同过滤算法只考虑相似度这一因素和传统矩阵分解出现负值的问题
提出了一种非负矩阵分解和项目热度相结合的两阶段k近邻选择算法
把项目热度融入到相似度计算中
有效缓解了数据稀疏性问题。实验结果表明:提出的算法与经典算法相比
推荐精度更高。
Abstract
Aiming at the problem that the neighbor-based collaborative filtering algorithm only considers the similarity factor and the negative value of the traditional matrix decomposition
a two-stage K-nearest neighbor selection algorithm combining non-negative matrix factorization and project heat is proposed. The project heat is integrated into the similarity calculation
the problem of data sparsity is effectively alleviated. Experimental results show that the proposed algorithm has higher precision than the classical algorithm.