Natural language processing and machine learning approaches for food categorization and nutrition quality prediction compared with traditional methods.

Our automation achieved high accuracy in classifying food categories and predicting nutrition quality scores using text information found on food labels. This approach is effective and generalizable in a dynamic food environment, where large amounts of food label data can be obtained from websites.