Probabilistic Rank and Reward: A Scalable Model for Slate Recommendation
This addresses scalable recommendation for low-latency domains like computational advertising, though it appears incremental as it builds on existing off-policy optimization methods.
The paper tackles the problem of personalized slate recommendation where users interact with at most one item, introducing Probabilistic Rank and Reward (PRR) to efficiently learn slate success probabilities by combining reward and rank information, resulting in outperforming existing off-policy methods and improved scalability to large action spaces.
We introduce Probabilistic Rank and Reward (PRR), a scalable probabilistic model for personalized slate recommendation. Our approach allows off-policy estimation of the reward in the scenario where the user interacts with at most one item from a slate of K items. We show that the probability of a slate being successful can be learned efficiently by combining the reward, whether the user successfully interacted with the slate, and the rank, the item that was selected within the slate. PRR outperforms existing off-policy reward optimizing methods and is far more scalable to large action spaces. Moreover, PRR allows fast delivery of recommendations powered by maximum inner product search (MIPS), making it suitable in low latency domains such as computational advertising.