The task of my master’s thesis was, after studying the related literature, to elaborate an approach using fuzzy rule based inference for the Artmedus smartphone application based patient support system, in order to present adequate contents helping pregnant users.
During this work I discovered a critical difficulty, whose solution could provide important aid for numerous pregnant users. This is the recommendation of suggested actions for pregnant users when certain symptoms are noticed. It is based on the degree of risks indicated by the symptoms.
As the solution for the problem I created a fuzzy model, which can be utilized for performing fuzzy inference. The construction of this model included the clustering of categorical samples, the definition of input and output dimensions as well as the establishment of a fuzzy rule base.
In order to validate the model conceptually, I built a software prototype, and using this prototype I performed simulations with the help of random training and test samples generated by myself. Finally, I evaluated the results.