Abstract—The recommendation system which recommend the interesting information for the target user must keep the high value of the precision and recall. However there is a trade-off relationship between precision and recall. In this paper, we propose the recommendation method keeping both precision and recall. The proposed method extracts the indicator from not only interesting information but also uninteresting information for the target user in the user-preferred-genre. Proposed method can keep the precision and recall by exclude the uninteresting information based on extracted indicator. From result of an experiment, we have verified the proposed method can improve the precision and recall.

Index Terms—recommendation system, uninterest indicator, precision, recall, trade-off

Cite:Tsukasa Kondo, Fumiko Harada, and Hiromitsu Shimakawa, "Recommendation System Keeping Both Precision and Recall Based on Uninteresting Information," Journal of Automation and Control Engineering, Vol.1, No.3, pp. 202-207, Sep. 2013. doi: 10.12720/joace.1.3.202-207