%0 Unpublished work
%T On food, bias and seasons: A recipe for sustainability
%+ École polytechnique (X)
%A Deudon, Michel
%8 2020-04-04
%D 2020
%Z Computer Science [cs]/Computation and Language [cs.CL]
%Z Life Sciences [q-bio]/Food and Nutrition
%Z Environmental Sciences/Environmental and Society
%Z Statistics [stat]/Machine Learning [stat.ML]Preprints, Working Papers, ...
%X Food is a common thread, linking all seventeen Sustainable Development Goals set by the United Nations (2016) for 2030. In this paper, we consider local-seasonal food as a proxy for social and environmental impact. We present a static and a dynamic generative model to re-sample ingredients from a dataset of 10k vegan recipes, in various context (location, season). We compare the static and dynamic behaviors in terms of greenhouse gas emissions and our results suggest that eating local-seasonal could save 0.25 to 1.5 kg CO2 per kg of product compared to randomly picked recipes, in Paris. We introduce a label, local-seasonal, to inform Human and Machine decisions for food and to protect/celebrate (bio)diversity. We propose an application to gather and share knowledge on local-seasonal food, worldwide, with professional and amateur cooks, farmers or markets, accessible at https://www.local-seasonal.org. We encourage initiatives to grow and support local communities as part of our recipe for sustainability.
%G English
%2 https://hal.archives-ouvertes.fr/hal-02532348/document
%2 https://hal.archives-ouvertes.fr/hal-02532348/file/VO.pdf
%L hal-02532348
%U https://hal.archives-ouvertes.fr/hal-02532348
%~ SDE
%~ X-DEP
%~ FRANTIQ
%~ GIP-BE
%~ X
%~ IP_PARIS
%~ IP_PARIS_COPIE