Orest Xherija


2022

pdf bib
CompLx@SMM4H’22: In-domain pretrained language models for detection of adverse drug reaction mentions in English tweets
Orest Xherija | Hojoon Choi
Proceedings of The Seventh Workshop on Social Media Mining for Health Applications, Workshop & Shared Task

The paper describes the system that team CompLx developed for sub-task 1a of the Social Media Mining for Health 2022 (#SMM4H) Shared Task. We finetune a RoBERTa model, a pretrained, transformer-based language model, on a provided dataset to classify English tweets for mentions of Adverse Drug Reactions (ADRs), i.e. negative side effects related to medication intake. With only a simple finetuning, our approach achieves competitive results, significantly outperforming the average score across submitted systems. We make the model checkpoints and code publicly available. We also create a web application to provide a user-friendly, readily accessible interface for anyone interested in exploring the model’s capabilities.

2018

pdf bib
Classification of Medication-Related Tweets Using Stacked Bidirectional LSTMs with Context-Aware Attention
Orest Xherija
Proceedings of the 2018 EMNLP Workshop SMM4H: The 3rd Social Media Mining for Health Applications Workshop & Shared Task

This paper describes the system that team UChicagoCompLx developed for the 2018 Social Media Mining for Health Applications (SMM4H) Shared Task. We use a variant of the Message-level Sentiment Analysis (MSA) model of (Baziotis et al., 2017), a word-level stacked bidirectional Long Short-Term Memory (LSTM) network equipped with attention, to classify medication-related tweets in the four subtasks of the SMM4H Shared Task. Without any subtask-specific tuning, the model is able to achieve competitive results across all subtasks. We make the datasets, model weights, and code publicly available.
Search
Co-authors
Venues