摘要
针对当前高校科研管理实际,研究对比了多种主流数据挖掘推荐算法的适用性,挑选出适合设备数据条件的推荐算法,并进行算法实际使用分析。最后,将基于内容过滤的推荐算法、基于用户的协同过滤的推荐算法和基于条目的 Slope One算法结合使用,互相补充,实现算法各性能的提高,完成高质量的推荐。
Considering the practical applicability of university scientific research management, the applicability of several mainstream data min- ing recommendation algorithms were compared in this paper. We choose suitable algorithm for the device data condition, and analyze actual us- ability of this alorithm. Finally, we achieve the improvement of the performance of algorithm and complete the high quality recommendations by the combination of recommendation algorithm based on content filtering, recommendation algorithm based on user collaborative and algorithm based on item slope one.
出处
《微型机与应用》
2016年第16期16-19,共4页
Microcomputer & Its Applications
关键词
推荐算法
数据挖掘
设备推荐
recommendation algorthm
data mining
device recommendation