Situation-dependent combination of long-term and session-based preferences in group recommendations: an experimental analysis.
SAC 2018: Symposium on Applied Computing Pau France April, 2018(2018)
摘要
A major challenge for conversational group recommender systems is how to properly exploit the user's preferences induced by the interactions between group members, which may deviate from the user's long-term ones. We argue that the relative importance of the long-term and group-induced preferences should vary according to the specific group settings. To support this claim, we compare alternative ways of combining these two types of user's preferences, under three diverse settings that may influence users' behavior. Our experimental study shows how a combination scheme that weighs more the long-term preferences can better serve the group when the group discussion has no impact on group members' preferences, but when the group context pushes members to have either more or less similar preferences, then users benefit from a recommender that weighs more the group-induced preferences.
更多查看译文
关键词
Recommender systems, group recommendations, conversational systems, preference elicitation
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络