Focus on what matters: Applying Discourse Coherence Theory to Cross Document Coreference

TL;DR

This paper introduces a discourse coherence theory-inspired method for cross-document coreference resolution, modeling reader focus to efficiently identify coreferent mentions, achieving state-of-the-art results across multiple datasets.

Contribution

It proposes a novel neighborhood-based embedding approach that constrains candidate mentions, enabling effective cross-document coreference resolution with improved robustness.

Findings

Achieves state-of-the-art results on multiple coreference datasets.

Training on multiple corpora improves average F1 by 17.2 points.

Model effectively handles inter-cluster coreference in cross-document scenarios.

Abstract

Performing event and entity coreference resolution across documents vastly increases the number of candidate mentions, making it intractable to do the full $n^2$ pairwise comparisons. Existing approaches simplify by considering coreference only within document clusters, but this fails to handle inter-cluster coreference, common in many applications. As a result cross-document coreference algorithms are rarely applied to downstream tasks. We draw on an insight from discourse coherence theory: potential coreferences are constrained by the reader's discourse focus. We model the entities/events in a reader's focus as a neighborhood within a learned latent embedding space which minimizes the distance between mentions and the centroids of their gold coreference clusters. We then use these neighborhoods to sample only hard negatives to train a fine-grained classifier on mention pairs and their local discourse features. Our approach achieves state-of-the-art results for both events and entities on the ECB+, Gun Violence, Football Coreference, and Cross-Domain Cross-Document Coreference corpora. Furthermore, training on multiple corpora improves average performance across all datasets by 17.2 F1 points, leading to a robust coreference resolution model for use in downstream tasks where link distribution is unknown.