江村, 裕介

EMURA, Yusuke

Lecture Recommend

協調フィルタリング

情報推薦

授業推薦

ユークリッド距離

Collaborative Filtering

Information Recommend

[Abstract] The Lecture Recommendation System has been developed to offer to meeting with "interesting lectures" which the students have not known based on the data of the Faculty of Engineering Chiba University. The Collaborative Filtering of the Lecture Recommendation System is utilized for the prevention of forgetting to register important lectures, and for the efficiency to choice lectures. In evaluation of the proposed system, the valid number of the recommended lectures is confirmed with F-measure in information retrieval, and the users of 90% answered that "interesting lectures" had been discovered in the questionnaire for the system. While the preference agreement ratio is increasing according to the recommendation ranking, the ratio is lower than the theoretical value at the top of the ranking, which is the problem in requirement of our proposed system.