Making Document-Level Information Extraction Right for the Right Reasons

TL;DR

This paper addresses the challenge of ensuring document-level information extraction models make correct inferences for the right reasons, proposing a framework that improves evidence quality without sacrificing accuracy.

Contribution

It introduces a post-hoc evidence extraction method with regularization and evidence supervision to enhance model interpretability and correctness in complex text inference tasks.

Findings

Regularization improves evidence extraction quality.

Models maintain accuracy while becoming more interpretable.

Evidence supervision enhances plausibility of inferences.

Abstract

Document-level models for information extraction tasks like slot-filling are flexible: they can be applied to settings where information is not necessarily localized in a single sentence. For example, key features of a diagnosis in a radiology report may not be explicitly stated in one place, but nevertheless can be inferred from parts of the report's text. However, these models can easily learn spurious correlations between labels and irrelevant information. This work studies how to ensure that these models make correct inferences from complex text and make those inferences in an auditable way: beyond just being right, are these models "right for the right reasons?" We experiment with post-hoc evidence extraction in a predict-select-verify framework using feature attribution techniques. We show that regularization with small amounts of evidence supervision during training can substantially improve the quality of extracted evidence. We evaluate on two domains: a small-scale labeled dataset of brain MRI reports and a large-scale modified version of DocRED (Yao et al., 2019) and show that models' plausibility can be improved with no loss in accuracy.