Inductive Pairwise Ranking: Going Beyond the n log(n) Barrier
This addresses the sample efficiency challenge in ranking tasks for applications like recommendation systems, though it builds incrementally on existing models.
The paper tackles the problem of ranking items from pairwise preferences using feature information, proposing the Feature Low Rank (FLR) model and the Inductive Pairwise Ranking (IPR) algorithm, which achieves good rankings with as little as 10% sampling on real-world datasets.
We study the problem of ranking a set of items from nonactively chosen pairwise preferences where each item has feature information with it. We propose and characterize a very broad class of preference matrices giving rise to the Feature Low Rank (FLR) model, which subsumes several models ranging from the classic Bradley-Terry-Luce (BTL) (Bradley and Terry 1952) and Thurstone (Thurstone 1927) models to the recently proposed blade-chest (Chen and Joachims 2016) and generic low-rank preference (Rajkumar and Agarwal 2016) models. We use the technique of matrix completion in the presence of side information to develop the Inductive Pairwise Ranking (IPR) algorithm that provably learns a good ranking under the FLR model, in a sample-efficient manner. In practice, through systematic synthetic simulations, we confirm our theoretical findings regarding improvements in the sample complexity due to the use of feature information. Moreover, on popular real-world preference learning datasets, with as less as 10% sampling of the pairwise comparisons, our method recovers a good ranking.