사용자의 다양성 성향을 고려하는 추천 기법

초록

Recently, a significant amount of research efforts have been devoted to diversification in recommendation systems. Previous approaches, however, are not satisfactory because they do not consider the diversity tendency of an individual user. To address this problem, we propose an approach that quantifies the degree of each user's diversity tendency through analyzing each user's previous preferences. Then, we propose a novel approach that uses the quantified factor in recommendation. Through a series of experiments, we demonstrate the superiority of our approach compared with existing ones in terms of both accuracy and diversity.

키워드