Stress Testing BERT Anaphora Resolution Models for Reaction Extraction in Chemical Patents
This work addresses the challenge of robust information extraction for chemical patent analysis, though it is incremental in focusing on noise robustness for a specific domain.
The researchers investigated how anaphora resolution models for extracting chemical reactions from patents perform in noise-free versus noisy environments, finding that model performance degraded significantly with noise but could be improved through targeted training strategies.
The high volume of published chemical patents and the importance of a timely acquisition of their information gives rise to automating information extraction from chemical patents. Anaphora resolution is an important component of comprehensive information extraction, and is critical for extracting reactions. In chemical patents, there are five anaphoric relations of interest: co-reference, transformed, reaction associated, work up, and contained. Our goal is to investigate how the performance of anaphora resolution models for reaction texts in chemical patents differs in a noise-free and noisy environment and to what extent we can improve the robustness against noise of the model.