基于蜂群K-means聚类模型的协同过滤推荐算法

针对目前协同过滤推荐算法的推荐质量和推荐效率低的问题,提出了一种基于改进蜂群K-means聚类模型的协同过滤推荐算法。首先,根据用户属性信息,采用改进蜂群K-means算法对用户进行聚类,建立用户聚类模型;然后,计算目标用户与用户聚类模型中各聚类中心的距离,其中距离最近的类为目标用户的检索空间;最后,从检索空间中依据用户-项目评分矩阵通过相似度计算搜索目标用户的最近邻居,由最近邻居的信息产生推荐列表。实验结果表明,该算法降低了平均绝对误差值,缩短了运行时间,提高了推荐质量和推荐效率。

A collaborative filtering recommendation algorithm based on a bee colony K-means clustering model

To address the problem of low recommendation quality and low recommendation efficiency of current collaborative filtering recommendation algorithms, we propose a collaborative filtering recommendation algorithm based on an improved bee colony K-means clustering model. Firstly, based on user attribute information, the algorithm uses the improved bee colony K-means algorithm to cluster users and establish a user clustering model. Secondly, we calculate the distance between target users and the clustering center in the user clustering model, and the cluster with the minimal distance is taken as the retrieval space of active users. Finally, we search the nearest neighbor of the target user by the similarity calculation according to the user item scoring matrix, and generate a recommendation list via the information of the nearest neighbor. Experimental results show that the proposed algorithm can achieve lower MAE and shorter running time, and it can enhance the quality and efficiency of recommendation.