摘要
传统电视点播VOD的协同过滤推荐算法被广泛应用,但随着大数据时代的到来,该算法的扩展性缺点也逐渐突出,传统算法越来越不适合解决海量数据带来的相似度伸缩性问题。本文提出一种基于hadoop框架的改进的协同过滤推荐算法,以弥补其缺陷。实验结果显示本算法的结果平滑性受海量相同视频估值问题影响相对较小,推荐准确性也有所优化。
Traditional VOD collaborative filtering recommendation algorithm is widely used,but with the advent of big data era,scalability shortcomings of the algorithm has become increasingly prominent,and traditional algorithm is not suitable for solving similarity scalability problem caused by massive data.This paper proposes an improved collaborative filtering recommendation algorithm based on hadoop framework to make up for its defects.The experimental results show that the smoothness of proposed algorithm is less affected by massive same video estimation problem,and recommendation accuracy is also optimized.
作者
余杰
李一凡
Yu Jie;Li Yifan(Jiangsu Cable Data Network Co.,Ltd,Jiangsu 210000,China)
出处
《广播与电视技术》
2021年第8期76-78,共3页
Radio & TV Broadcast Engineering
关键词
协同过滤
扩展性问题
大数据
Collaborative filtering
Scalability
Big data
