Overview: Drugs can potentially lead to adverse reactions or side effects, and having prior knowledge of these reactions is crucial for preventing hospitalizations and premature deaths. Existing public databases for common adverse drug reactions (ADRs) primarily rely on individual reports from drug manufacturers and healthcare professionals. However, this passive approach to ADR surveillance often suffers from significant under-reporting. In response to this challenge, I recognize the value of social media platforms, including online health forums, where patients worldwide voluntarily share their experiences with drug intake. These platforms offer a vast and untapped source of information such as askapatient.com, webmd.com, and iodine.com, twitter.com and other sources for identifying unreported ADRs.
Solution: In this GitHub repository, I will build the ADR Detection Framework. This framework leverages Natural Language Processing (NLP) techniques to automatically identify and extract ADRs from drug reviews sourced from social media platforms.
Project Key Features: -> Collects and processes drug-related data from various social media sources. -> Utilize state-of-the-art NLP algorithms for ADR detection. -> Provides tools and resources for ADR data analysis and visualization. -> Offer a scalable and adaptable solution for ADR surveillance.