I am trying to calculate coherence score in topic modeling.

I am following this Github link

So there I need to use the preprocessed wiki and news. I got 3 questions:

  1. if the domain that I have applied my topic modeling is different from wiki and news does it still make sense to use them for finding the co-ocurance and finally coherence?

  2. the final score has been calculated for each topic. for example lets say I apply coherence on 20 news group data set and set number of topic to be 20. In the papers they normally report one coherence score for the whole 20 topics, however here it assigns coherence for each topic separately. Does it make sense to average them and report one score here?

  3. the link provided in the read me does not work for downloading pre-oroceesed wiki and news, I was wondering if any of you guys have a workable link for those sources?


Your Answer

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

Browse other questions tagged or ask your own question.