“…A major challenge with recipe datasets is that no standard online collections exist yet due to restrictions in terms of use, language, etc. Previous research often uses proprietary materials through APIs including crawls from online recipe collections and databases such as epicurious.com , allrecipes.com , RecipeDB, CulinaryDB, Hawa World, TarlaDalal.com , and chefkoch.de ( Shidochi et al, 2009 ; Ahn et al, 2011 ; Teng et al, 2012 ; Ahnert, 2013 ; Jain et al, 2015 ; Kusmierczyk et al, 2015 ; Bogojeska et al, 2016 ; Tallab and Alrazgan, 2016 ; Kikuchi et al, 2017 ; Sajadmanesh et al, 2017 ; Bagler and Singh, 2018 ; Chang et al, 2018 ; Min et al, 2018 ; Asano and Biermann, 2019 ; Batra et al, 2019 ; Trattner and Elsweiler, 2019 ; Herrera, 2020 ; Sharma et al, 2020 ). To be useful in practical applications, these untapped sources require structuring, linking, and analysis via NLP techniques.…”