Hi!
As it is not possible to run the Topic Modeling script on Extracted terms (while it is possible to run the same script on extracted Named Entities…), I am looking for a way to circumvent the problem. That’s why I wonder: does it make sense to upload a corpus, extract terms, download the results, reupload the extracted terms only (not the whole texts) as a corpus and run the TM script on it? The real question is: does TM need text (with sentences etc) or a list of terms by document is enough?

I am not talking about the technical feasibility but whether it does make sense or not.
Thanks
fb