I settled on a regex based approach with lots of data clean-up and normalisation up front. Example of my site [4] Some other approaches I spent a lot of time o

I settled on a regex based approach with lots of data clean-up and normalisation up front. Example of my site [4] Some other approaches I spent a lot of time on: * Extracting Structured Data From Recipes Using Conditional Random Fields [1] * Chef Watson [2] * Ingredient Parser - Model Guide [3] [1] https://archive.nytimes.com/open.blogs.nytimes.com/2015/04/09/extracting-structured-data-from-recipes-using-conditional-random-fields/?_r=0 https://archive.nytimes.com/open.blogs.nytimes.com/2015/04/0... [2] https://blog.kitchenpc.com/2011/07/06/chef-watson/ https://blog.kitchenpc.com/2011/07/06/chef-watson/ [3] https://ingredient-parser.readthedocs.io/en/latest/guide/index.html https://ingredient-parser.readthedocs.io/en/latest/guide/ind... [4] https://pretty-recip.es/recipe?recipe-url=https%3A%2F%2Fwww.seriouseats.com%2Fchard-pancetta-pecorino-quiche-recipe https://pretty-recip.es/recipe?recipe-url=https%3A%2F%2Fwww....