A Case-Based Solution to the Cold-Start Problem in Group Recommenders

Abstract

Abstract. We extend a group recommender system with a case base of previous group recommendation events. We show that this offers a potential solution to the cold-start problem. Suppose a group recommendation is sought but one of the group members is a new user who has few item ratings. We can copy ratings into this user's profile from the profile of the most similar user in the most similar group from the case base. In other words, we copy ratings from a user who played a similar role in some previous group event. We show that copying in this way, i.e. conditioned on groups, is superior to copying nothing and also superior to copying ratings from the most similar user known to the system.

Case-Based Aggregation of Preferences for Group Recommenders

“…The second observation is that results with Soc are better than results with Std : the use of the social information improves the quality of the recommendations. This is not a new result [11,8]. But what is new, our third observation, is the performance of the CBR system.…”

“…For the comparison, we use a group recommender that we developed previously [11,8]. With real data and, in more recent work, with a larger dataset of artificial data, we showed that, relative to simpler approaches, our group recommender improves the accuracy of predicted group ratings and the precision of group recommendations, and that is why we use it here.…”

“…We leave this to future work. We denote the group similarity by gsim cbr , and we emphasize that this is a new definition, richer than definitions that we have used in other work [8]. In effect, it is a form of graph similarity: users are nodes; trust relationships are weighted edges.…”

“…Hence, we created our own dataset. We have explained its construction elsewhere [8], and so we only summarize here.…”

Performance Analysis of Group Recommendation Systems in TV Domains

“…Researchers evaluate and verify group recommendation approaches, however, with synthetic or limited datasets [8][9][10][11][12][13][14][15][16] because real large-scale datasets of group profiles are rare. Regarding design group recommendation systems, we need to consider group and domain characteristics simultaneously.…”