Research of User-Based Collaborative Filtering Recommendation Algorithm Based on Hadoop

DOI

10.2991/cisia-15.2015.17How to use a DOI?

Keywords

collaborative filtering; similarity measure; recommended system; hadoop

Abstract

Collaborative filtering algorithm is one of the key technologies of the current e-commerce recommendation system, in which the effect of similarity measure directly determines the accuracy of the recommendation system. An improved method of similarity measure and the corresponding collaborative filtering recommendation algorithm by introducing a common user items popularity rating between features and user relevance is proposed in this paper. Furthermore, the implementation of collaborative filtering recommendation system based on hadoop is discussed. The respective evaluation of traditional collaborative filtering recommendation algorithm and improved recommendation algorithm by using MAE show that the improved algorithm enhances the recommendation accuracy in a certain extent. Meanwhile, the overall performance experiments show that collaborative filtering recommendation engines continue to reduce the calculation time with the appropriate increase of the virtual machine node.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Y.L. Zhang
AU  - M.M Ma
AU  - S.P Wang
PY  - 2015/06
DA  - 2015/06
TI  - Research of User-Based Collaborative Filtering Recommendation Algorithm Based on Hadoop
BT  - Proceedings of the International Conference on Computer Information Systems and Industrial Applications
PB  - Atlantis Press
SP  - 63
EP  - 66
SN  - 2352-538X
UR  - https://doi.org/10.2991/cisia-15.2015.17
DO  - 10.2991/cisia-15.2015.17
ID  - Zhang2015/06
ER  -