Speakers: Linjia Chang and Lav R. Varshney, University of Illinois at Urbana-Champaign
Abstract: Recent research in computational gastronomy has been focusing on using data of recipes, ingredients, and chemical compounds to create interesting dishes with pleasant flavor combinations. The flavor pairing hypothesis has been tested among European, Asian, and Arab cuisines. However, texture, one of the key sensory properties in determining people’s perception for food, is rarely studied on a network level. The only related texture research in the food industry is mainly for designing individual commodities.
We collected a set of data of 30k recipes each with 17 texture scores and applied network analysis to study the texture pairing patterns of recipes in 11 different regions around the world. The result reveals that different regions have different texture pairing patterns, among which some are preferred globally while others are rarely practiced anywhere. We also show preliminary results of texture construction of recipes from ingredients.
(This work is in collaboration with Vivanda.)