Biomedical relation extraction with pre-trained language representations and minimal task-specific architecture

TL;DR

This paper demonstrates that a minimal architecture built on pre-trained BERT can effectively extract gene-disease relation triples in biomedical texts, outperforming baselines despite class imbalance.

Contribution

The study shows that fine-tuning BERT with a simple classification layer suffices for biomedical relation extraction, requiring minimal task-specific engineering.

Findings

Outperforms random baseline in relation extraction

Effective despite class imbalance

Uses minimal task-specific architecture

Abstract

This paper presents our participation in the AGAC Track from the 2019 BioNLP Open Shared Tasks. We provide a solution for Task 3, which aims to extract "gene - function change - disease" triples, where "gene" and "disease" are mentions of particular genes and diseases respectively and "function change" is one of four pre-defined relationship types. Our system extends BERT (Devlin et al., 2018), a state-of-the-art language model, which learns contextual language representations from a large unlabelled corpus and whose parameters can be fine-tuned to solve specific tasks with minimal additional architecture. We encode the pair of mentions and their textual context as two consecutive sequences in BERT, separated by a special symbol. We then use a single linear layer to classify their relationship into five classes (four pre-defined, as well as 'no relation'). Despite considerable class imbalance, our system significantly outperforms a random baseline while relying on an extremely simple setup with no specially engineered features.