Link to dataset: https://www.kaggle.com/hugodarwood/epirecipes#full_format_recipes.json
Recipe recommendation: given a list of ingredients, output a list of recommended recipes. Dataset used: expiration_data, foodList, recipe
Fuzzy word matching: given a list of OCR input strings (including shopping mall, location, dates, and purchased items), output all the food items found in the strings. The strings are filtered and matched twice against the 6k entry foodList and the more comprehensive 200k row food_categ. The algorithm is supported by Python's fuzzywuzzy library.