An NLP-Based Medicine Recommendation System Using Bidirectional GRU for Disease Prediction

Abstract: The fast progress of Artificial Intelligence in healthcare has created chances for smart disease prediction and automatic medicine suggestion systems. Old ways of diagnosing diseases often need a lot of expertise from humans, time and medical resources which're not always available in far away or poor areas. To solve these problems this research suggests a system that uses Natural Language Processing to recommend medicine. This system uses a kind of model called Bidirectional Gated Recurrent Unit for predicting diseases and suggesting medicine accurately. This system uses Natural Language Processing to prepare and analyze the symptoms that patients have. It uses methods like breaking down text into parts, cleaning making all texts the same length and giving labels to codes to change raw medical text into something that machines can read. The Bidirectional model is used to understand how symptoms are related to each other in both directions, which helps to understand medical patterns better. The model is trained on a dataset that has kinds of diseases and is balanced. The results of the experiment show that this approach is than 98 percent accurate and has strong precision, recall and F1-score values. This system is also less complicated than models and is very efficient and can be used by many people. The system that is suggested can be used for healthcare, telemedicine and clinical assistance systems that use Artificial Intelligence. In the future it may be possible to use this system on the cloud support languages and connect it to devices that monitor healthcare in real time. Artificial Intelligence, in healthcare is getting better and better. This system can be a part of it. The fast progress of Artificial Intelligence is helping to make disease prediction and medicine suggestion systems.

Keywords: Natural Language Processing, Bidirectional GRU, Disease Prediction, Medicine Recommendation, Deep Learning, Healthcare Artificial Intelligence, Symptom Analysis.