Shah, D.1, Ahn, R. J.1 and Cohen, J. M.1, (1)Analytical Flavor Systems, New York, NY, USA
Poster
The Gastrograph system is a sensory platform that enables panelists to describe the product they are tasting across 24 flavor attributes and somatosensations supplemented by the specific reference flavors the panelists taste in any specific product. At the end of each review, the panelists are asked to assign a perceived quality score for that product. Most sensory panels at breweries do not contain a stratified sampling of the general population, so standard statistical methods cannot be employed in order to understand the preferences of the average beer consumer. To project the preferences of professional panelists onto the preferences of the general population, reviews that were completed on the Gastrograph system were sampled from, in accordance with the tasting experience level distribution of the general population. The techniques LFDA (local Fisher discriminant analysis) and PAM (partitioning around medoids) were used to maximize between-product similarity and minimize within-product similarity in flavor profile. The random forest method is then utilized to predict the distribution of perceived quality scores the general population would assign given any set of reviews, with built-in considerations for the class of beer and the tasting experience of the panelists.
Before starting Analytical Flavor Systems, Jason Cohen was the founder and executive director of The Tea Institute at Penn State, which oversees 20+ researchers in 5 fields of study in traditional Chinese, Japanese, and Korean teas. Jason did his research in sensory science and data mining, eventually developing the Gastrograph system after three and a half years of research. Jason is a professional coffee, tea, and beer taster.
View Presentation