基于聚类算法的数字图书馆知识推送原理

Research on knowledge pushing service of digital library based on clustering algorithms

文章针对高校师生用户对数字图书馆的推送服务满意度进行了调研分析,结果显示用户对图书馆的推送服务满意程度不高。基于此类问题,文章根据高校用户属性特点进行k-means聚类研究,使推荐信息时考虑用户自身属性特点,包括年纪、专业和目的等特点,并据此设计了混合属性的距离函数。建立的基于聚类算法的数字图书馆知识推送服务,提高了原有相似信息对用户的模糊推荐效果,提高了用户体验。

According to the survey and analysis about the push service satisfaction of the university teachers and students’ users to the digital library, the results show that the users are not satisfied with the push service. Based on this kind of problems, according to the characteristics of university users’ attributes, K-means clustering is studied, so that users’ own attributes, including age, specialty, and purpose, are taken into account when pushing information, and a distance function of mixed attributes is designed in this paper. Based on clustering algorithm, the knowledge push service of digital library improves the blurred recommendation effect of original similar information to users and improves the users’ experience.