I'm currently working on a topic model in R for the purpose of extracting topics from groups of tasks. I've been using an LDA with the topicmodels package, and while this runs, it's not exactly what I'm looking for in terms of topic coherence. I've tried separating these tasks to one document each, and while this has yielded slightly better results, I would like to explore options that could produce even better results. I recently read a paper on the effectiveness of a GSDMM on extracting topics from short-form text documents, and I want to pursue this type of topic model in R.
I am fairly new to topic modeling in R, and after some research into GSDMM, I haven't been able to find a lot of literature on doing one in R.
I am just curious if this is possible in R, and if so, what would that look like? I know that this topic modeling technique is more complex to do compared to an LDA, but I would just like some advice on where to look for info.
Like I had mentioned, I've run an LDA and it produces results, but I would like my topics to be less related to each other. Breaking down these tasks and lowering the alpha value did yield better results.