0

I am working on a system that compares a source document to a target document and then generate alternative variations of the source document. The goal is to reach a higher cosine similarity between the two documents, say +80%. There's already plenty of articles on how to find the similarity difference, but not many on how to generate suggestions or improvements.

I've never heard of a model that can do this, so how do I build a model that can suggest altered variations of a source document with the ultimate goal of increasing its similarity to a given target document?

Articles on document similarity

Samewise
  • 1
  • 1
  • Linking the few articles that you found would be nice, and maybe also adding some info about what you're planning to try or tried already. As it is phrased right now the question sounds like asking other people to do your job – Edoardo Guerriero Feb 07 '22 at 12:32
  • I don't really understand the problem. If you have the target document, is your ideal source document the target document itself (Since it has 1 cosine similarity)? – DKDK Feb 08 '22 at 02:39

0 Answers0