A Machine Learning Driven Malaria Fever Medical Consultation Chatbot

Abstract

Nearly half of the world's population is at risk of malaria fever. There were reported 247 million cases of malaria fever as at the year 2021 with 619000 deaths. This shows malaria is yet a serious burden to the world. Patients now have additional options for receiving medical information and services as a result of the increased use of chatbot systems in the healthcare delivery. In this work, a Machine Learning technique -Decision Tree was trained using malaria dataset collected from Adetoyin Hospital, Ado-Ekiti, Nigeria. The decision rules from the decision tree were then used to train the medical chatbot using Feed Forward Neural Network (FNN) Natural Language Processing (NLP) technique. Performance evaluation results of both the Decision Tree model and the medical chatbot were encouraging. The developed system is intended to discuss normally with patients and obtain pertinent medical history and symptoms of malaria fever and thereafter offers medical diagnosis. This experiment shows medical chatbot systems have the potential to improve patient access to healthcare services by giving patients a quick, easy, and personalized source of medical advice; these technologies can aid in the digital revolution of healthcare.