lda-project / lda

Topic modeling with latent Dirichlet allocation using Gibbs sampling

Home Page:https://lda.readthedocs.io/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Is it possible to get topic assignments of words?

HaoboGu opened this issue · comments

Thanks for your great work!

I checked the documentation, it seems that I could only get counts of topic-word assignments(nzw_).

So is it possible to get topic assignments of words?

Do you mean model.topic_word_?

Thanks for the reply!
But what I want to know is the specific topic assigned to each word, rather than topic-word distribution.

lda doesn't keep precise track of word order so there's no explicit tracking of
topic assignment to specific words.

You can, however, sample the topic assignment for an individual word
given knowledge of the posterior distribution.

On 04/25, JasonGu wrote:

Thanks for the reply!
But what I want to know is the specific topic assigned to each word, rather than topic-word distribution.


You are receiving this because you commented.
Reply to this email directly or view it on GitHub:
https://github.com/ariddell/lda/issues/55#issuecomment-214293779

Got it.

Thanks very much :)