[Topic-models] LDA for collaborative filtering

Feng Yan (严丰) fengyan0 at gmail.com
Wed Nov 19 05:06:44 EST 2008


Hi Everyone,

I'm investigating use LDA for collaborative filtering tasks. In Blei's
original paper, the authors map vocabulary to movie, document to user.
However, as we know a document consists of a bag of words, multiple
occurrences of the same words is possible. So my question is how many
occurrences do you use for each movie?

For example, a user rate movie A 3-star, movie B 2-star, what's the user's
corresponding document? Is it "A B" or "A A A B B"?

-- 
Feng Yan (fengyan0 at gmail.com)
Undergraduate Student
Department of Mathematical Sciences, Tsinghua University, Beijing, P.R.China
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cs.princeton.edu/pipermail/topic-models/attachments/20081119/6705caa4/attachment.html>


More information about the Topic-models mailing list