Hi, @dongwookim-ml
I am going to use the RTM model, however, I am not able to understand the datatype. Can you provide what datatype RTM expects, so that I would be able to use this model on my own data?

my concern is to add new documents out of the training set, then after natural processing applying relational topic model to measure the similarity between these documents.
is there any examples that help me in my issue

Hi arongdari，
I meet a problem when use your source code named “at_model.py” . In the seventh line， from six.moves import xrange ,

So could you tell me what wrong with the code ?
Many thanks!

It should be lambda_v instead of lambda_u at initialization of matrix V(line 50).

I have to add some regularizer term to the likelihood equation containing beta terms. I am unable to figure out from the code which part of the code does the M-step update for the conditional multinomial parameter beta

doc_cnt : structure and what is it?
doc_ids: is it a 1-D array of all the doc ids?
n_voca?

n_topics: I believe items are the actual documents. So, if I have 1000 documents, n_topics will be equal to 1000

Hi, maybe you should provide a way to set the loglevel when creating a model. Or an additional getter/setter pair. Thanks in advance.

Can you explain how to create the list giving the links between documents for the RTM model.
In the RTM example notebook it was imported directly without any explanation and I can't figure out how to create it for a new dataset
Thanks

Thanks for a great package -- I just got the author-topic model successfully running and I was wondering whether there is a simple way to get the per-document topic distribution for (a) the documents an author-topic model was fitted on or (b) new documents. Thanks in advance for any replies!

Requires to make a test code for each model.

Hi I am trying to implement LDA in tensorflow. I am quite new to both tensorflow and LDA. Currently I am following your lda_vb implementation.
Is to possible to have a vectorized implemenation (without for loops) of do_e_step method?
If yes, it would very helpful if you provide some insights on how to implement it.

In the notebooks in python-topic-model/notebook/, there are no small examples provided of how to infer the topic distribution for a new document or for the documents that the model was trained on.

Something like giving a list of integers as input (that map to the words of voca) for a new document, and getting the probability distribution that this document has for the trained topics. Or accessing the topics of all the trained documents.

How can this be achieved for lets say the LDA or the supervised LDA?

Hi,

For author-topic model, could you please provide an example of showing topic of docs after model being trained?
I find only ways to show topic distribution of author and word distribution of topic, however, in my case, I care topic of docs much more.

Thanks.

I tried running the author-topic model notebook. I noticed that the topic distributions of many authors were flat, meaning that all topics were equally likely. See example below.

I did not change the notebook in any way, so I suspect there is some error in the algorithm/code, although I have no inkling of what it might be. Just thought I'd share.

Is someone else able to reproduce this, or is it just me? Or did I misunderstand, and this is actually expected to happen?

Hi Dongwoo:
I am currently looking for a gibbs sampling estimation method for supervised LDA, your Stochastic (Gibbs) EM for sLDA (slda_gibbs.py) is exactly what i'm looking for.
I was wondering if there're any papers or other materials that can explain the math behind it, especially the matrix calculation part?

Many thanks!