Supporting User Interaction in Flavor Sampling Trials

Sensory innovation in the flavor industry is a process that relies heavily on user involvement to carry out critical sampling trials at a high cost. We are interested in how recommender systems, in particular, have an important role to play in view of addressing some of their key challenges and constraints. We describe a purpose-built sensory recommendation prototype with the potential to: (1) help users navigate a vast and complex solution space using very limited user preference information, (2) reduce the time it takes to gather critical information about individual user flavor preferences, and at the same time (3) allow for a broader coverage of the flavor space to be evaluated during inhouse flavor development trials, thus potentially reducing related overheads (e.g., time and money). In addition, we summarize some of the findings following a real-user evaluation of the prototype recently carried out by a leading Flavor industrial partner.