Combining Reward and Rank Signals for Slate Recommendation - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2021

Combining Reward and Rank Signals for Slate Recommendation

Imad Aouali
  • Fonction : Auteur
Sergey Ivanov
  • Fonction : Auteur
Mike Gartrell
  • Fonction : Auteur
David Rohde
  • Fonction : Auteur
Flavian Vasile
  • Fonction : Auteur
Victor Zaytsev
  • Fonction : Auteur
Diego Legrand
  • Fonction : Auteur

Résumé

We consider the problem of slate recommendation, where the recommender system presents a user with a collection or slate composed of K recommended items at once. If the user finds the recommended items appealing then the user may click and the recommender system receives some feedback. Two pieces of information are available to the recommender system: was the slate clicked? (the reward), and if the slate was clicked, which item was clicked? (rank). In this paper, we formulate several Bayesian models that incorporate the reward signal (Reward model), the rank signal (Rank model), or both (Full model), for non-personalized slate recommendation. In our experiments, we analyze performance gains of the Full model and show that it achieves significantly lower error as the number of products in the catalog grows or as the slate size increases.
Fichier principal
Vignette du fichier
aouali_combining.pdf (701.94 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03959628 , version 1 (27-01-2023)

Identifiants

Citer

Imad Aouali, Sergey Ivanov, Mike Gartrell, David Rohde, Flavian Vasile, et al.. Combining Reward and Rank Signals for Slate Recommendation. KDD '21 Workshop on Bayesian Causal Inference for Real World Interactive Systems, Aug 2021, Virtual Event, Singapore. ⟨hal-03959628⟩
11 Consultations
28 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More