Individualized extreme dominance (IndED): A new preference-based method for multi-objective recommender systems

Reinaldo Silva Fortes, Daniel Xavier de Sousa, Dayanne Gouveia Coelho, Anísio Mendes Lacerda, Marcos André Gonçalves

Abstract: Highlights•We propose IndeD, a new preference-based method for multi-objective recommendation.•A new dominance relation concept considering the users’ preferences is defined.•The new decision making process minimizes the distance to the user’s preferences.•Our method beats competitive baselines in meeting individualized users’ preferences.•IndED obtains the best results in the optimization of the most difficult objectives.