Medical assistance bot

African Institute for Mathematical Sciences,Mbour,Senegal | Senegal

Foutse Yuehgoh | Dieu Merci KIMPOLO NKOKOLO | Mohamed Cheik Ibrahim TRAORE | Barnabas Kouami

The fact that health care system in Africa is in an abominable state is no surprise for anybody. This is particularly true of the doctor-to-population ratio and the number of hospitals. As of 2004, according to the World Bank, the average health workforce per 100,000 (one hundred thousand) people in Sub-Saharan Africa is 15.5. Even today this has not changed a lot. One of the consequences is that physicians are overwhelmed by the number of patients. Their efficiency diminishes with tiredness. As for the patients, they have to queue for a long time before being able to see a physician. Also some people are far from any kind of medical center. These make people turn to self-medication because of the accessibility of health care.We want to build a mobile application that will allow users to get diagnosis and prescriptions for their disease by discussing with a chatbot. The user will have the opportunity to find the nearest pharmacy. Also the platform will allow for health awareness campaigns (epidemic, vaccination campaign, ...). As chatbots do not have any specific goal that guides the interaction, the difficulty of evaluation is intrinsic as each conversation is interactive, and the same conversation will not occur more than once; one slightly different answer will lead to a completely different conversation; moreover there is no clear sense of when such a conversation is “complete”. More still, identifying an appropriate database is a challenge as well. Fortunately, chatbots have recently become the focus of greater research interest. As such many solutions are already available to mimic some challenges significantly. All these allows us to confidently believe in this idea.