基于聚类优化的协同过滤个性化图书推荐

Personalized Book Recommendation Based on Cluster Optimization and Collaborative Filtering

图书借阅是图书馆最基本的服务,根据用户的借阅爱好为其自动地推荐相关图书是解决图书借阅效率与可靠性等问题的关键。为了提高图书推荐的准确率,本文利用改进的K-mean算法对借阅用户的类别与偏好性进行了系统的分析,然后通过构造用户借阅偏好性矩阵与用户相似性度量,采用协同过滤算法实现了图书借阅的个性化推荐。实验结果表明,本文算法可根据用户的借阅爱好准确地为其推荐图书,整体上具有较高的性能。

Book lending is the fundamental library service, and automatically recommending appropriate books for users according to their reading preferences is the key to solve the book lending problems in both efficiency and reliability. In order to improve the accuracy of book lending, the paper utilizes an improved K-mean algorithm to systematically analyze the categories and the reading preferences of users, and then constructs the reading preference matrix and user similarity measures to perform the personalized recommendation of book lending by using the collaborative filtering algorithm. Experimental results show our proposed method can accurately recommend books for users, and has high performance.