How to categorize ingredients in a recipe's ingredients list?

Question

Python

Good afternoon!
Student. Specialty - programmer. Familiar with Python 3.0. I have no experience with neural networks. Left without a leader, but with the approval of the theme of the project. We need a scheme of the methodological part, which we could really implement.
Idea. The ingredients of a recipe are divided into groups: Containing mainly (i) carbohydrates - potatoes, flour; (ii) proteins - meat, cottage cheese; (iii) fats - butter, sunflower oil. Task. Teach the network to reliably identify in a recipe (in the list of ingredients of a recipe) an ingredient with a predominant content of carbohydrates, proteins, etc. Input data. Since the work is in the department of natural language processing, at the input are individual ingredients in the form of text (the name of the ingredient), at the output is the group to which the ingredient belongs - carbohydrate, protein, fat. The names need to be converted into vectors and the neural network must be trained to determine the type of ingredient. Save. I will be glad for any hint.

Answer 1

Well, here neyruha is unlikely to be able to guess where to put the avocado, because it is impossible to establish a connection between the letters in the name of the product and bzhu. You can make a reference book, and use it to give out which group this or that ingredient belongs to, but not neurukha.