Text Classification for Effective Question Answering By Chatbot

Abstract

This paper offers a novel framework for creating an advanced chatbot that uses the Llama-2 model to answer medical questions. The chatbot, which is intended to provide all-encompassing medical support in multiple domains, makes use of the sophisticated natural language processing (NLP) features built into the Llama-2 architecture. The dataset that was used comes from "The Gale Encyclopedia of Medicine 2," and it has been carefully pre-processed to make sure it works with the experimental setup. Cleaning, formatting, and text extraction are important preprocessing tasks. By fine tuning the Llama-2 model on the medical dataset, the methodology allows it to be customized to accurately read and reply to user inquiries in the healthcare domain. Notably, the chatbot's cutting-edge capabilities improve user experience and medical accuracy. These include individualized responses, regional language processing, and the distribution of safety precautions. Thorough analysis reveals the chatbot's exceptional 85% symptom diagnostic accuracy and effectiveness in providing tailored medical support. This research incorporates cutting-edge features and modifications catered to user requirements, going beyond conventional medical question-answering systems. In order to increase accessibility and patient outcomes, future research areas might focus on developing the chatbot's capabilities even further and investigating how to incorporate it into larger healthcare projects