Computer Engineering and Applications ›› 2022, Vol. 58 ›› Issue (5): 104-111.DOI: 10.3778/j.issn.1002-8331.2101-0387

• Big Data and Cloud Computing • Previous Articles     Next Articles

Hybrid Recommendation Algorithm Combining Wolf Colony Algorithm and Fuzzy Clustering

WANG Yonggui, LI Xin   

  1. Software College, Liaoning Technical University, Huludao, Liaoning 125105, China
  • Online:2022-03-01 Published:2022-03-01

融合狼群算法和模糊聚类的混合推荐算法

王永贵,李昕   

  1. 辽宁工程技术大学 软件学院,辽宁 葫芦岛 125105

Abstract: Aiming at the data sparsity problem of traditional collaborative filtering recommendation algorithm and the limitation of searching for similar users, a hybrid recommendation algorithm combining wolf colony algorithm and fuzzy clustering is proposed. Firstly, in the process of data processing, according to the project-based collaborative filtering algorithm, the data relationship between projects is fully mined, and the zero value of the original matrix is filled to reduce the data sparsity; secondly, from the user’s point of view, according to the size of the membership degree of fuzzy clustering, the relevant neighbor set is selected to expand the search scope of relevant users; the wolf colony algorithm is introduced into fuzzy clustering, with the help of fuzzy clustering, wolf colony algorithm has the advantage of global search to improve the accuracy of finding similar users. The experimental results on real datasets show that the proposed algorithm alleviates the problem of data sparsity, reduces the recommendation error significantly, and has a good recommendation effect compared with the traditional recommendation algorithm.

Key words: collaborative filtering, fuzzy C-means clustering, wolf colony algorithm, hybrid recommendation

摘要: 针对传统协同过滤推荐算法普遍存在的数据稀疏性问题以及寻找相似用户时存在局限性,提出一种融合狼群算法和模糊聚类的混合推荐算法。在数据处理过程中,根据基于项目的协同过滤算法充分挖掘项目间的数据关系,填充原始矩阵的零值以降低数据稀疏性;从用户的角度出发,根据模糊聚类隶属度的大小筛选出相关邻居集合,扩大相关用户的寻找范围;将狼群算法引入模糊聚类,借助狼群算法全局搜索的优势,提高寻找相似用户的准确度。在真实的数据集上进行对比实验,结果表明,所提算法缓解了数据稀疏的问题,推荐误差明显减小,和传统的推荐算法相比有着良好的推荐效果。

关键词: 协同过滤, 模糊C-均值聚类, 狼群算法, 混合推荐