A Framework for Optimizing Paper Matching

TL;DR

This paper introduces a framework that optimizes the assignment of papers to reviewers in scientific conferences by learning suitability scores and solving an integer programming problem, improving matching efficiency.

Contribution

It presents a novel framework combining learning and optimization for paper-reviewer matching, reducing reviewer burden and enhancing assignment quality.

Findings

Learning methods improve suitability score accuracy

Matching formulations outperform baseline methods

Framework reduces reviewer workload

Abstract

At the heart of many scientific conferences is the problem of matching submitted papers to suitable reviewers. Arriving at a good assignment is a major and important challenge for any conference organizer. In this paper we propose a framework to optimize paper-to-reviewer assignments. Our framework uses suitability scores to measure pairwise affinity between papers and reviewers. We show how learning can be used to infer suitability scores from a small set of provided scores, thereby reducing the burden on reviewers and organizers. We frame the assignment problem as an integer program and propose several variations for the paper-to-reviewer matching domain. We also explore how learning and matching interact. Experiments on two conference data sets examine the performance of several learning methods as well as the effectiveness of the matching formulations.