Clarify that Bertopic classify documents when LDA extract topics

[emilienschultz]

After Béatrice presentation and a comment from Carlo, I think it could be interesting to explicitely say that Bertopic and LDA pursue 2 different goals :

• LDA : extract topics and gives mixture of topics
• Bertopic : classify document

And maybe add that it is possible to get closer to LDA (as an application)

• sentence granurality for the document
• create topics
• agregate proportions at the document scale