Graph Neural Re-Ranking via Corpus Graph

TL;DR

This paper introduces Graph Neural Re-Ranking (GNRR), a novel approach that leverages graph neural networks to incorporate document relationships into re-ranking, significantly improving ranking performance by considering broader document context.

Contribution

The paper proposes a new GNN-based re-ranking pipeline that models document relationships via corpus subgraphs, enhancing query-document relevance evaluation.

Findings

GNRR improves ranking metrics on TREC-DL19 by 5.8% in Average Precision.

GNNs effectively capture cross-document interactions for better re-ranking.

Incorporating document context via graphs enhances re-ranking performance.

Abstract

Re-ranking systems aim to reorder an initial list of documents to satisfy better the information needs associated with a user-provided query. Modern re-rankers predominantly rely on neural network models, which have proven highly effective in representing samples from various modalities. However, these models typically evaluate query-document pairs in isolation, neglecting the underlying document distribution that could enhance the quality of the re-ranked list. To address this limitation, we propose Graph Neural Re-Ranking (GNRR), a pipeline based on Graph Neural Networks (GNNs), that enables each query to consider documents distribution during inference. Our approach models document relationships through corpus subgraphs and encodes their representations using GNNs. Through extensive experiments, we demonstrate that GNNs effectively capture cross-document interactions, improving performance on popular ranking metrics. In TREC-DL19, we observe a relative improvement of 5.8% in Average Precision compared to our baseline. These findings suggest that integrating the GNN segment offers significant advantages, especially in scenarios where understanding the broader context of documents is crucial.