Cool Beans! TSYS School Researchers Develop Expert Coffee Rating System

In their research study appearing in IEEE Transactions on Fuzzy Systems, the TSYS School’s Javier Livio and Rania Hodhod describe "cupping" is the process of sensorial evaluation, using a numerical score, of coffee beans, which is done to determine the actual sensory differences between coffee samples, to describe the flavors of the samples, and to determine preference of product.  Cupping is typically done by expert judges who record individual components’ scores and rank their intensities for reference. 

In their study, Livio and Hodhod introduce a fuzzy expert (logic) system that linguistically models a cupper’s perception of coffee bean attributes through sensorial evaluation and is capable of learning as new cupping data flow through it.  Based on tests indicating that the fuzzy system matches experts’ ratings at a 95 percent rate, the researchers assert that their fuzzy expert system is capable of training expert cuppers, especially in cases where coffees from several countries are being evaluated.