Difference between revisions of "Yano et al NAACL 2009"
m |
|||
| Line 12: | Line 12: | ||
This paper expands upon LinkLDA, presented in [[RelatedPaper::Erosheva 2004 Mixed membership models of scientific publications|Erosheva et al. (2004)]]. | This paper expands upon LinkLDA, presented in [[RelatedPaper::Erosheva 2004 Mixed membership models of scientific publications|Erosheva et al. (2004)]]. | ||
| + | Here, <math>\theta</math> is a distribution over topics, <math>\beta</math> is a multinomial distribution over post words, and <math>\gamma</math> is a multinomial distribution over (comment) users. <math>N</math> and <math>M</math> are the words counts in the post and all of its comments, respectively. | ||
[[Image:link_LDA.png|250px]] | [[Image:link_LDA.png|250px]] | ||
Although LinkLDA can model which users are likely to respond to a post, it does not model the comment text they will write. | Although LinkLDA can model which users are likely to respond to a post, it does not model the comment text they will write. | ||
| − | The authors expand on this by proposing CommentLDA. | + | The authors expand on this by proposing CommentLDA, as shown below. |
[[Image:comment_LDA.png|300px]] | [[Image:comment_LDA.png|300px]] | ||
| + | |||
| + | In CommentLDA, note that the comment text is modeled by the distribution over comment words given topics, <math>\beta^{\prime}</math>. | ||
== Experimental Result == | == Experimental Result == | ||
Revision as of 09:45, 26 September 2012
Contents
Citation
Tae Yano, William Cohen, and Noah A. Smith. Predicting Response to Political Blog Posts with Topic Models. In Proc of NAACL 2009.
Online Version
Predicting Response to Political Blog Posts with Topic Models.
Summary
This Paper describes a topic model based approach in modeling the generation of blog text (posts and comments).
Brief description of the method
This paper expands upon LinkLDA, presented in Erosheva et al. (2004). Here, is a distribution over topics, is a multinomial distribution over post words, and is a multinomial distribution over (comment) users. and are the words counts in the post and all of its comments, respectively.
Although LinkLDA can model which users are likely to respond to a post, it does not model the comment text they will write. The authors expand on this by proposing CommentLDA, as shown below.
In CommentLDA, note that the comment text is modeled by the distribution over comment words given topics, .
Experimental Result
Task: given a training dataset consisting of a collection of blog posts and their commenters and comments, and a unseen test dataset from a later time period, predict who is going to comment on a new blog post from the test set.
Dataset is available at
The compared models were:
