NeurIPS 2020

Learning Rich Rankings

Meta Review

The paper proposes a new "CRS" model for ranking, provides theoretical guarantees for MLE (which also apply to MNL as a special case), and real-world experiments. Given the very nice conceptual, theoretical and empirical results, this paper is recommended for acceptance. The paper has the wrong style file for the rebuttal. While the rebuttal is not ignored for this reason since the rebuttal at least obeys the spirit of the style file, request the authors to be much more careful in the future.