5

I am needing to develop a product taxonomy for foods and beverages that would work across multiple food providers. I need this to provide a code hierarchy for mapping specific products in a categorical fashion, e.g.

Large Latte -IS-A-> Coffee Beverage -IS-A-> Beverage

Double espresso -IS-A-> Coffee Beverage -IS-A-> Beverage

The only thing that I've found so far is a USDA Food Group dictionary that has some high level categories.

So, to be clear, in response to the confusion I seemingly have caused others trying to help, I'm looking for a code-dictionary that provides hierarchical (taxonomic) categories of food products typically found in restaurants.

  • I'm not sure I follow the "->". Are those categorical inclusions? I assume you've looked at dbpedia.org and wikidata.org? – Barry Carter Aug 6 '16 at 23:51
  • Thanks, Barry. The arrows do indeed represent categorical inclusions - ontologically they would be is_a directed edges. The domain is quite tight, in that I am looking only for food products that would be found in typicall restaurants, and where I would label the leaves with closest parent nodes. Ontology is the highest form of result here, whereas even a taxonomical mapping would prove useful. Think of these categories as "food products I can buy in the covered restaurants". – Shawn Mehan Aug 7 '16 at 0:00
  • OK, I just edited your question to make it clearer that you had two rows there, instead of one row, and also corrected the spelling in one row. dbpedia.org may help: dbpedia.org/page/Latte and dbpedia.org/page/Espresso though not double espresso. I haven't looked at wikidata for this specific question but it's fairly good too. – Barry Carter Aug 7 '16 at 0:33
  • have you looked at the lives spec for health inspections? it may not get this in depth, but sounds kind of similar to what you want. – albert Aug 8 '16 at 4:14
2

I'm not aware of an open data ready-made taxonomy that's tuned for restaurants, but you can start with nutrient databases (like the Canadian one: http://www.hc-sc.gc.ca/fn-an/nutrition/fiche-nutri-data/index-eng.php) and reverse engineer the 5,690 foods they describe into a taxonomy. If you can consider a commercial dataset, John Snow Labs (www.johnsnowlabs.com) have built a food taxonomy for mobile health apps, so it covers both 'common' meals and restaurant foods.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.