PharmaPlus - GenerativeAI Chatbot for Streamlined Drug Informations

Abstract

This is a BERT-based Large Language Model (LLM) powered chatbot intended to provide all the necessary information regarding Humira, a medication used for treating autoimmune diseases. The chatbot simplifies the intricate medical details into clear and actionable insights to assist both the healthcare professionals and patients. The system preprocesses the Humira dataset using a structured natural language pro-cessing pipeline, which includes dependency parsing, stemming, lemmatization, stopword removal, sentence segmentation, and word tokenization. The polished text is broken down into smaller chunks, put through the Word2Vec model for conversion to numerical embeddings, and saved in a vector database to provide faster retrieval. The system returns relevant passages when a query is submitted, which are then run through the BERT model to generate contextual answers. A Flask-based web interface is leveraged to display these answers in a very user-friendly manner, ensuring smooth interaction. This lowers prescription mistakes, promotes informed healthcare choices, and increases access to reliable pharmacologic information by combining state-of-the-art approaches in NLP with AI-driven models.

Ähnliche Arbeiten