Crowdsourcing-Based Knowledge Graph Construction for Drug Side Effects Using Large Language Models with an Application on Semaglutide

2025-04-06

AI Summary

This study addresses the insufficient detection of real-world adverse drug reaction (ADR) signals in pharmacovigilance. We propose a novel knowledge graph construction paradigm integrating crowdsourced data with large language models (LLMs). Methodologically, we extract patient-reported narratives on semaglutide (for weight management) from social media platforms (e.g., Reddit), employ LLMs for entity recognition, relation extraction, and temporal modeling of unstructured text, and perform cross-source alignment and validation against the FDA Adverse Event Reporting System (FAERS). Key contributions include: (1) the first comprehensive ADR knowledge graph dedicated to multiple branded formulations of semaglutide; (2) identification of 12 temporally distinct safety signals—78% of which are corroborated by FAERS; and (3) substantial enhancement of dynamic pharmacoepidemiological insights and knowledge base completion derived from patient-generated narratives.

Technology Category

• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Machine Learning: Large Multimodal Models (LMMs)
• Multiagent Systems: Adversarial Agents

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Agentic search

Abstract

Social media is a rich source of real-world data that captures valuable patient experience information for pharmacovigilance. However, mining data from unstructured and noisy social media content remains a challenging task. We present a systematic framework that leverages large language models (LLMs) to extract medication side effects from social media and organize them into a knowledge graph (KG). We apply this framework to semaglutide for weight loss using data from Reddit. Using the constructed knowledge graph, we perform comprehensive analyses to investigate reported side effects across different semaglutide brands over time. These findings are further validated through comparison with adverse events reported in the FAERS database, providing important patient-centered insights into semaglutide's side effects that complement its safety profile and current knowledge base of semaglutide for both healthcare professionals and patients. Our work demonstrates the feasibility of using LLMs to transform social media data into structured KGs for pharmacovigilance.

Problem

• Extracting drug side effects from noisy social media data
• Constructing knowledge graphs using large language models
• Validating findings against pharmacovigilance databases

Innovation

• Leveraging LLMs to extract drug side effects
• Organizing social media data into knowledge graphs
• Validating findings with FAERS database comparisons