Item-Based Clustering CF Recommend Algorithm in MapReduce Framework

DOI

10.2991/icacie-16.2016.6How to use a DOI?

Keywords

clustering, collaborative filtering, recommend algorithm, mapreduce, hadoop

Abstract

Due to the overload of information, it's difficult to find the information we need from large amount of information. Thus, the recommendation algorithm appears. When the size of data is too large, the traditional collaborative filtering recommendation algorithm becomes unable to satisfy the requirement, such as out of memory. Therefore, this paper proposed a collaborative filtering algorithm based on feature clustering. Using the co-occurrence matrix replaces the similarity matrix and enhancing the efficiency of the algorithm. The item-based collaborative filtering algorithm was improved in order to make the algorithm run in parallel modes, such as running on Hadoop. So the algorithm can be easily improved performance by adding more nodes. After that, I clustered the rating data by the feature of user's, generated intra-group and inter-group recommendation results and mixed the two recommendation results. The experimental platform is Hadoop. And experiments show that: the complexity of matrix calculation has been reduced while the accuracy of recommendation results has been improved.

Copyright

© 2016, 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  - Liquan Han
AU  - Yuqiang Jiang
AU  - Qi Chu
PY  - 2016/10
DA  - 2016/10
TI  - Item-Based Clustering CF Recommend Algorithm in MapReduce Framework
BT  - Proceedings of the 2016 International Conference on Automatic Control and Information Engineering
PB  - Atlantis Press
SP  - 24
EP  - 27
SN  - 2352-5401
UR  - https://doi.org/10.2991/icacie-16.2016.6
DO  - 10.2991/icacie-16.2016.6
ID  - Han2016/10
ER  -