Abstract
In this paper, a voice-based disease prediction system, has been developed to extract symptoms, status, and possible treatments by assessing performance on common symptoms. The designed system uses a decision tree algorithm for prediction of disease based on the patient symptoms. The outcome of the research is focused to blend in with the traditional methods to the machine learning model in a way that works in converting patient speech to text in the application and to describe the most basic class of inspection in medication. In addition, the proposed methodology is limited to run on only voice samples provided and the model as designed can be extended to introduce more robust version to support the users offering natural language interface and connecting the users to specialists in the field of concern.