Want to Identify, Extract and Normalize Adverse Drug Reactions in Tweets? Use RoBERTa

TL;DR

This paper utilizes RoBERTa to effectively identify, extract, and normalize adverse drug reactions in tweets, significantly outperforming average scores in shared tasks related to health social media mining.

Contribution

The paper introduces a RoBERTa-based system for ADR detection, extraction, and normalization in social media, achieving notable improvements over average scores in shared tasks.

Findings

F1-score of 58% in ADR classification, 12% above average

Relaxed F1-score of 70.1% in ADR extraction, 13.7% above average

Relaxed F1-score of 35% in ADR normalization, 5.8% above average

Abstract

This paper presents our approach for task 2 and task 3 of Social Media Mining for Health (SMM4H) 2020 shared tasks. In task 2, we have to differentiate adverse drug reaction (ADR) tweets from nonADR tweets and is treated as binary classification. Task3 involves extracting ADR mentions and then mapping them to MedDRA codes. Extracting ADR mentions is treated as sequence labeling and normalizing ADR mentions is treated as multi-class classification. Our system is based on pre-trained language model RoBERTa and it achieves a) F1-score of 58% in task2 which is 12% more than the average score b) relaxed F1-score of 70.1% in ADR extraction of task 3 which is 13.7% more than the average score and relaxed F1-score of 35% in ADR extraction + normalization of task3 which is 5.8% more than the average score. Overall, our models achieve promising results in both the tasks with significant improvements over average scores.