基于人口统计学的改进聚类模型协同过滤算法

摘要

针对传统基于用户的协同过滤推荐算法在大数据环境下存在评分高维稀疏性、推荐精度低的问题,提出一种基于人口统计学数据与改进聚类模型相结合的协同过滤推荐算法,以提高推荐系统精度和泛化能力.该方法首先通过用户人口统计学数据属性,结合用户-项目评分矩阵计算各个用户间的相似度;然后对用户、项目进行分层近邻传播聚类,根据用户对项目的评分数据计算用户或项目之间的相似性,产生目标用户或项目的兴趣近邻;最后根据兴趣最近邻进行推荐.对Epinions,MovieLents等数据集进行仿真实验,仿真的结果表明,与传统的协同过滤算法相比,提出的算法提高了推荐精度,为传统的协同过滤推荐算法提供了参考.%The traditional user based collaborative filtering recommendation algorithm in large data environment has the problem of high dimensional sparse and low recommendation accuracy.A collaborative filtering recommendation algorithm based on the combination of demographic data and improved clustering model was proposed to improve the accuracy and generalization ability of the recommendation system.Firstly,this method calculates the similarity among different users through the user demographic data attributes and the user-item score matrix.Secondly,hierarchical neighbor clustering of user and project,calculates the similarity between users or items by the user's score data for the project,and generates interest in a neighbor of a target user or project.Finally,according to the recent interest in the nearest neighbor to recommend.Simulation experiments on Epinions and MovieLents data set,the simulation results show that the proposed algorithm improves the recommendation accuracy compared with the traditional collaborative filtering algorithm,provide reference for the traditional collaborative filtering recommendation algorithm.