I have a large dataset of short reviews and I would like to find the most recurring themes. For this reason, I got into topic modeling.
I am looking for some good tutorials and references for short text topic modeling. I have seen that there exists a lot of material for LDA but when it comes to short texts there is not much, especially for python implementations. I have found this StackOverflow issue but is more than 2 year and a half old.
So, I would like to know if something more recent exists. I know that there is a transformer-based model called BertTopic that should perform reasonably well even on short text. Let me know if there is something that I am missing.
Thank you!