ART: group recommendation approaches for automatically detected groups

摘要: Group recommender systems provide suggestions when more than a person is involved in the recommendation process. A particular context which group useful number of lists that can be generated limited (i.e., it not possible to suggest list items each user). In such case, grouping users and producing recommendations groups becomes necessary. None approaches literature able automatically order overcome previously presented limitation. This paper presents set detect by clustering them, respect constraint on maximum produced. The proposed have been largely evaluated two real-world datasets compared with hundreds experiments statistical tests, validate results. Moreover, we introduce best practices help development this context.