TY - GEN
T1 - A user modeling using implicit feedback for effective recommender system
AU - Oh, Jehwan
AU - Lee, Seunghwa
AU - Lee, Eunseok
PY - 2008
Y1 - 2008
N2 - An amount of information on the Web has been increased explosively with the growth of information technology. In the area of electronic commerce, the recommender systems that provide personalized content are crucial research issue. The analysis of efficient user preference is important for improving the recommendation accuracy. Existing recommendation system has used implicit feedback for analyzing user preference. But, when the collected user information is lack, it is not fit. This paper proposes a personalized recommendation system which is based on information built by analyzing implicit feedback. The proposed system monitors the various user behaviors comprehensively to analyze user intention more precisely. The system also deduces the most important attribute for the user among various attribute of a product based on ID3 algorithm, and applies the result to analyzing user preference. Therefore, proposed system is able to recommend items in the situation that user behavior information is lack. Empirical results show that the effectiveness of the system is confirmed.
AB - An amount of information on the Web has been increased explosively with the growth of information technology. In the area of electronic commerce, the recommender systems that provide personalized content are crucial research issue. The analysis of efficient user preference is important for improving the recommendation accuracy. Existing recommendation system has used implicit feedback for analyzing user preference. But, when the collected user information is lack, it is not fit. This paper proposes a personalized recommendation system which is based on information built by analyzing implicit feedback. The proposed system monitors the various user behaviors comprehensively to analyze user intention more precisely. The system also deduces the most important attribute for the user among various attribute of a product based on ID3 algorithm, and applies the result to analyzing user preference. Therefore, proposed system is able to recommend items in the situation that user behavior information is lack. Empirical results show that the effectiveness of the system is confirmed.
UR - https://www.scopus.com/pages/publications/55849143655
U2 - 10.1109/ICHIT.2008.274
DO - 10.1109/ICHIT.2008.274
M3 - Conference contribution
AN - SCOPUS:55849143655
SN - 9780769533285
T3 - Proceedings - 2008 International Conference on Convergence and Hybrid Information Technology, ICHIT 2008
SP - 155
EP - 158
BT - Proceedings - 2008 International Conference on Convergence and Hybrid Information Technology, ICHIT 2008
T2 - 2008 International Conference on Convergence and Hybrid Information Technology, ICHIT 2008
Y2 - 28 August 2008 through 29 August 2008
ER -