Combining Reward and Rank Signals for Slate Recommendation

TL;DR

This paper introduces Bayesian models for slate recommendation that integrate reward and rank signals, demonstrating improved accuracy especially with larger catalogs and slate sizes.

Contribution

It proposes a novel Bayesian framework combining reward and rank signals for non-personalized slate recommendation, showing performance improvements.

Findings

Full model achieves lower error than models using only reward or rank signals.

Performance gains are more significant as catalog size increases.

The combined model scales well with larger slate sizes.

Abstract

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.