Sarker Lab Emory University

Prompt Collections

Reusable prompt templates developed by our lab for biomedical NLP tasks.

RAG-Based Dynamic Prompting for Biomedical NER

Few-shot biomedical named entity recognition using retrieval-augmented generation with dynamic prompting.

Automated Thematic Analysis of Health Content

Automating inductive thematic analyses of health-related social media content using LLMs.

HILGEN: NER Data Generation with Knowledge Bases

Hierarchically-informed data generation for biomedical NER using knowledge bases and LLMs.

Healthcare Text Classification

Benchmarking open-source LLMs on healthcare text classification tasks including adverse drug events and health mentions.

Medical Error Detection and Correction

Knowledge-enhanced few-shot in-context learning for detecting and correcting errors in clinical text.

Blood Pressure Extraction from Literature

Extracting blood pressure values and demographic breakdowns from scientific literature using LLMs.

ADHD Detection from Narrative Transcripts

Leveraging LLMs for ADHD detection from naturalistic movie viewing narrative transcripts.

Xylazine Wound Management Thematic Analysis

Automated thematic analysis of xylazine wound management discussions from social media.

Health-Related Social Media Text Classification

Evaluating LLMs as zero-shot classifiers for health-related text classification on social media.