I am new to NLP and NLTK in particular. I have a list of products such as: shampoo, toothpaste, vitamins, brushes, soap, etc. 1- or 4-grams. What I am trying to do is to walk several levels up the taxonomy tree and output a set of categories, like bath products, hair styling, diet supplements, etc.
My thoughts that it can be implemented via similarity computation between individual products and if it's above some threshold (0.5?), navigate hypernyms until you get a match.
Maybe there is an easier way?