I will check out Watson. I'm really just looking to tag ingredients. I could always create my own tag standards for the what I'm curious about.

I have read through this thread to try to make myself more articulate. I guess I'm thinking of it as an ingredient centered approach. The recipe is secondary to whatever is available. The idea would be the ingredients would have tags for properties. The properties would be the nutritional content, cultural references, mouth feel, taste profile, allergens, ...

The combining of ingredients might have less to do with recipes and more to do with necessities and circumstances.

The flour example is probably the most illustrative. Many grains and roots can be used as flour. If one was using taro root flour all recipes would just accommodate that substitution.

This is an interesting and complicated set of problems. It's similar to problems in the LLM & social network spaces. You have several dimensions, like recipes, ingredients, nutrition, cultural context, etc, and you have gradients to help determine recommendations, which points to a graph DB. But you also have rules when it comes to substitutions (not every substitution is appropriate for every situation), and you might have rules on other dimensions as well. That points to a layer of business logic. But to serve recommendations performantly, you might want to run rules upon writes or on a schedule to persist the results to the DB. Either that or do some sophisticated caching.

I knew I was asking in the right place. The good news is…

The idea is not commercial in nature

The system wouldn’t need to be exhaustive. The data that is already in databases would probably suffice to get started.

The system would be tailored and refined by individuals and groups.

There would be a mechanical turk aspect for graphing/validating “recipes”.

Thanks for commenting.