I'm building a food semantic search model and I want to use a pre-trained SentenceTransformers model with cosine similarity. I'm using Epicurious dataset for the corpus which consists of textual ("title", "description", "directions") as well as categorical ("categories") and numerical data ("calories", "fat", "sodium").

The model that computes embeddings for the corpus takes String data as an input. So my idea was to combine the categorical and numerical data with textual data in a single String.

Do you think it's a right way to handle categorical data? If yes, what are the ways to concatenate such data (maybe using column names with the values or adding some tags)? Otherwise, what is a better way to handle such data?



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.