TSPRank: Bridging Pairwise and Listwise Methods with a Bilinear Travelling Salesman Model

TL;DR

TSPRank introduces a novel hybrid ranking method that models the ranking task as a Travelling Salesman Problem, effectively combining pairwise and listwise approaches to improve global ranking performance across diverse applications.

Contribution

It proposes TSPRank, a new hybrid ranking approach that leverages combinatorial optimisation to enhance learning-to-rank models by integrating pairwise and listwise methods.

Findings

TSPRank significantly outperforms traditional pairwise and listwise methods.

It improves ranking performance across stock, retrieval, and historical data tasks.

The method effectively captures global information for better rankings.

Abstract

Traditional Learning-To-Rank (LETOR) approaches, including pairwise methods like RankNet and LambdaMART, often fall short by solely focusing on pairwise comparisons, leading to sub-optimal global rankings. Conversely, deep learning based listwise methods, while aiming to optimise entire lists, require complex tuning and yield only marginal improvements over robust pairwise models. To overcome these limitations, we introduce Travelling Salesman Problem Rank (TSPRank), a hybrid pairwise-listwise ranking method. TSPRank reframes the ranking problem as a Travelling Salesman Problem (TSP), a well-known combinatorial optimisation challenge that has been extensively studied for its numerous solution algorithms and applications. This approach enables the modelling of pairwise relationships and leverages combinatorial optimisation to determine the listwise ranking. This approach can be directly integrated as an additional component into embeddings generated by existing backbone models to enhance ranking performance. Our extensive experiments across three backbone models on diverse tasks, including stock ranking, information retrieval, and historical events ordering, demonstrate that TSPRank significantly outperforms both pure pairwise and listwise methods. Our qualitative analysis reveals that TSPRank's main advantage over existing methods is its ability to harness global information better while ranking. TSPRank's robustness and superior performance across different domains highlight its potential as a versatile and effective LETOR solution.