Tailored Recommendations
Résumé
Many popular internet platforms give personalized recommendations to their users, based on other users who have provided similar ratings for some items. We propose a novel approach to such recommendation systems by viewing a recommendation as a way to extend an agent's preferences to items she has not yet rated, by means of some aggregate of other agents' preferences. These extension and aggregation requirements are expressed by an Acceptance and a Pareto principle, respectively. We show through examples that so-called collaborative filtering systems used by popular platforms typically violate the Pareto principle. We then develop a formal model within which we identify the recommendation systems satisfying the above two principles. A central feature of this model is the use of incomplete preference relations to handle agents who have not rated all items.
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|
Loading...