Designing a chatbot to analyze patient sentiment : The case of the Sangarébougou Community Health Center

Abstract

This paper examines the integration of deep learning and natural language processing (NLP) to develop medical chatbots that provide diagnoses and recommendations based on user-provided symptoms. The rural commune of Sangarébougou in Mali serves as a case study for the development and deployment of this chatbot. The project begins with the collection and preprocessing of a corpus of health data, followed by the analysis and tokenisation of the textual data. Various NLP and deep learning algorithms are then applied to train the chatbot model. The implementation uses several Python libraries, including TensorFlow and NLTK, to run the text processing and machine learning models. The results show that modern NLP and deep learning techniques can achieve satisfactory performance in supporting automated medical. The developed chatbot can understand user questions and provide accurate, relevant answers, thereby enhancing healthcare accessibility in rural areas.