The dataset contains 100 short text entries, each about 150–200 characters, describing common food items from five cuisines. Each sample has two features: the text (a short description) and the cuisine (the categorical label to predict). To expand the data, 1,000 synthetic samples were created using simple augmentation methods such as synonym replacement, ingredient swaps, small rephrasings, and adding or removing minor… See the full description on the dataset page: https://huggingface.co/datasets/scottymcgee/food-text-dataset.
