Mention Annotations Alone Enable Efficient Domain Adaptation for Coreference Resolution

TL;DR

This paper presents a method for efficient domain adaptation in coreference resolution by annotating only mentions, significantly reducing annotation effort while improving model performance across diverse datasets.

Contribution

The authors introduce a mention-only annotation approach combined with a high-precision detection objective, enabling effective domain transfer with less annotation effort.

Findings

Achieves 7-14% F1 improvement across datasets

Mention annotation is nearly twice as fast as full coreference annotation

Method facilitates annotation-efficient transfer in diverse domains

Abstract

Although recent neural models for coreference resolution have led to substantial improvements on benchmark datasets, transferring these models to new target domains containing out-of-vocabulary spans and requiring differing annotation schemes remains challenging. Typical approaches involve continued training on annotated target-domain data, but obtaining annotations is costly and time-consuming. We show that annotating mentions alone is nearly twice as fast as annotating full coreference chains. Accordingly, we propose a method for efficiently adapting coreference models, which includes a high-precision mention detection objective and requires annotating only mentions in the target domain. Extensive evaluation across three English coreference datasets: CoNLL-2012 (news/conversation), i2b2/VA (medical notes), and previously unstudied child welfare notes, reveals that our approach facilitates annotation-efficient transfer and results in a 7-14% improvement in average F1 without increasing annotator time.