Adapting Coreference Resolution for Processing Violent Death Narratives

TL;DR

This paper enhances coreference resolution in violent death narratives, especially for LGBT-related texts, by developing data augmentation techniques to address domain-specific challenges and improve model transferability.

Contribution

It introduces data augmentation rules within a probabilistic framework to improve coreference models for gender-inclusive administrative narratives.

Findings

Data augmentation improves coreference resolution accuracy.

Enhanced models better handle LGBT-related narrative texts.

Demonstrated effectiveness on CDC violent death data.

Abstract

Coreference resolution is an important component in analyzing narrative text from administrative data (e.g., clinical or police sources). However, existing coreference models trained on general language corpora suffer from poor transferability due to domain gaps, especially when they are applied to gender-inclusive data with lesbian, gay, bisexual, and transgender (LGBT) individuals. In this paper, we analyzed the challenges of coreference resolution in an exemplary form of administrative text written in English: violent death narratives from the USA's Centers for Disease Control's (CDC) National Violent Death Reporting System. We developed a set of data augmentation rules to improve model performance using a probabilistic data programming framework. Experiments on narratives from an administrative database, as well as existing gender-inclusive coreference datasets, demonstrate the effectiveness of data augmentation in training coreference models that can better handle text data about LGBT individuals.