Fine-grained Information Status Classification Using Discourse Context-Aware BERT

TL;DR

This paper introduces a discourse context-aware BERT model that significantly improves fine-grained information status classification and bridging anaphora recognition, outperforming previous methods without relying on complex hand-crafted features.

Contribution

The paper presents a novel BERT-based approach for fine-grained information status classification that achieves state-of-the-art results and reduces dependence on manual linguistic features.

Findings

Achieves 4.8% accuracy improvement on IS classification.

Improves bridging anaphora recognition by 10.5 F1 points.

Model's attention signals align with linguistic notions of information status.

Abstract

Previous work on bridging anaphora recognition (Hou et al., 2013a) casts the problem as a subtask of learning fine-grained information status (IS). However, these systems heavily depend on many hand-crafted linguistic features. In this paper, we propose a simple discourse context-aware BERT model for fine-grained IS classification. On the ISNotes corpus (Markert et al., 2012), our model achieves new state-of-the-art performance on fine-grained IS classification, obtaining a 4.8 absolute overall accuracy improvement compared to Hou et al. (2013a). More importantly, we also show an improvement of 10.5 F1 points for bridging anaphora recognition without using any complex hand-crafted semantic features designed for capturing the bridging phenomenon. We further analyze the trained model and find that the most attended signals for each IS category correspond well to linguistic notions of information status.