FnTaste

FnTaste is a cross-platform open-source software library designed to model, analyze, and simulate human taste perceptions in culinary experiments. It provides a modular toolkit for representing taste attributes, aroma descriptors, texture, and user preferences, along with tools for recipe parsing and flavor profiling.

The project originated in the mid-2010s as an effort to formalize sensory data within computational gastronomy. It evolved through community-driven releases with emphasis on portability, interoperability with culinary databases, and extensible flavor lexicons. Key milestones include a standardized taste-scoring model and a plugin system for ingredient libraries.

FnTaste comprises core modules such as a data model for taste and aroma descriptors, a scoring engine that normalizes results across sessions, a recipe analysis module, and a visualization layer for sensory data. It supports external datasets and allows researchers to calibrate models using user studies, with an emphasis on reproducibility and open data practices.

Usage and reception: It is used in academic settings and in product development to explore flavor optimization, menu design, and educational demonstrations. Critics note that taste perception is inherently subjective, so computational results should complement human sensory evaluation rather than replace it.