融入改进的K-means聚类的协同过滤算法的研究与应用
2021-08-17刘鑫



摘 要:本文通过对K-means聚类算法和协同过滤推荐算法的学习研究。针对基于用户的协同过滤算法的不足,将改进的K-means聚类算法融入其中,设计了基于K-means聚类算法的个性化推荐算法,并将其应用于旅游景点及线路的个性化推荐中,以提高个性化推荐质量。实验结果表明,基于改进的K-means聚类的协同过滤算法缓解了初始数据的稀疏性问题,针对不同用户喜爱的旅游景点及线路推荐,在准确率和召回率两个方面证明可以提高个性化推荐的准确度。
关键词:K-means聚类;协同过滤算法;最小生成树
中图分类号:TP183 文献标识码:A DOI:10.3969/j.issn.1003-6970.2021.03.027
本文著录格式:刘鑫.融入改进的K-means聚类的协同过滤算法的研究与应用[J].软件,2021,42(03):097-099
Research and Application of Collaborative Filtering Algorithm Incorporating Improved K-means Clustering
LIU Xin
(Jilin Institute of Architecture and Technology, Changchun Jilin 130114)
【Abstract】:This article is based on the study of K-means clustering algorithm and collaborative filtering recommendation algorithm. Aiming at the deficiencies of the user-based collaborative filtering algorithm, the improved K-means clustering algorithm is incorporated into it, and a personalized recommendation algorithm based on the K-means clustering algorithm is designed and applied to the personalized recommendation of tourist attractions and routes In order to improve the quality of personalized recommendations. The experimental results show that the collaborative filtering algorithm based on improved K-means clustering alleviates the sparsity problem of the initial data. It is proved that it can improve the personalization in terms of accuracy and recall rate for different users' favorite tourist attractions and routes recommendation. Recommended accuracy.
【Key words】: K-means clustering;collaborative filtering algorithm;minimum spanning tree
0引言
随着互联网的飞速发展,信息资源呈现几何级数增长,用户在大量的信息中很难获取到真正需要的数据信息。传统的搜索引擎已无法满足用户的特殊需求,因此个性化推荐系统应运而生,并成为了解决信息过载问题的有效方法。一个完整的个性化推荐系统总体框架包含三个部分:数据特征提取模块、个性化推荐模块、推荐结果模块。个性化推荐系统能够根据用户个人喜好自動进行信息推荐,减少信息的冗余,简化用户的操作,因此个性化推荐算法也逐渐成为学术界的研究热点之一[1]。
1主要技术
1.1协同过滤算法及其存在的问题
推荐系统中协同过滤推荐算法是当前应用最为广泛,研究最多,影响最为深远的个性化推荐技术,协同过滤推荐算……
