Comments on: Modeling blog post comment counts http://livewebir.com/blog/2008/07/modeling-blog-post-comment-counts/ by Paul Ogilvie Mon, 15 Feb 2010 03:14:51 +0000 hourly 1 http://wordpress.org/?v=3.0.1 By: Michelle http://livewebir.com/blog/2008/07/modeling-blog-post-comment-counts/comment-page-1/#comment-7 Michelle Tue, 25 Nov 2008 11:28:21 +0000 http://pogil.wordpress.com/?p=8#comment-7 Hi, Paul, thanks for your reply and the updated details. Now I understand it better. I apologize for two typos, i.e., Y=X/(1+X) (should be Y=X/(1-X)) and X=Y/(1-Y) (should be X=Y/(1+Y)) in my earlier comment. Hi, Paul, thanks for your reply and the updated details. Now I understand it better.

I apologize for two typos, i.e., Y=X/(1+X) (should be Y=X/(1-X)) and X=Y/(1-Y) (should be X=Y/(1+Y)) in my earlier comment.

]]> By: Michelle http://livewebir.com/blog/2008/07/modeling-blog-post-comment-counts/comment-page-1/#comment-6 Michelle Sun, 23 Nov 2008 14:21:55 +0000 http://pogil.wordpress.com/?p=8#comment-6 Hi, this is a great post. I understand the part the posterior p|{x1,...,xn} was induced. However, what is not equivalently intuitive to me is the part about the posterior r|{x1,...,xn}. Could you talk a bit more about why beta prime distribution was chosen for r? As mentioned, ``if X ~ Beta(a, b), then Y = X/(1+X) ~ Beta'(a, b)''. In another word, if Y ~ Beta'(a, b), then X = Y/(1-Y) ~ Beta(a, b). If `Y' is substituted by r, which variable will be the substitution of `X'? Hi, this is a great post. I understand the part the posterior p|{x1,…,xn} was induced. However, what is not equivalently intuitive to me is the part about the posterior r|{x1,…,xn}. Could you talk a bit more about why beta prime distribution was chosen for r?

As mentioned, “if X ~ Beta(a, b), then Y = X/(1+X) ~ Beta’(a, b)”. In another word, if Y ~ Beta’(a, b), then X = Y/(1-Y) ~ Beta(a, b). If `Y’ is substituted by r, which variable will be the substitution of `X’?

]]> By: Why we model comment counts < Information Retrieval on the Live Web http://livewebir.com/blog/2008/07/modeling-blog-post-comment-counts/comment-page-1/#comment-3 Why we model comment counts < Information Retrieval on the Live Web Wed, 12 Nov 2008 05:36:18 +0000 http://pogil.wordpress.com/?p=8#comment-3 [...] Retrieval on the Live Web by Paul Ogilvie < Modeling blog post comment counts What makes a blog post popular? series at mSpoke blog [...] [...] Retrieval on the Live Web by Paul Ogilvie < Modeling blog post comment counts What makes a blog post popular? series at mSpoke blog [...]

]]> By: Why We Model Comment Counts « Paul Ogilvie http://livewebir.com/blog/2008/07/modeling-blog-post-comment-counts/comment-page-1/#comment-2 Why We Model Comment Counts « Paul Ogilvie Tue, 08 Jul 2008 19:34:04 +0000 http://pogil.wordpress.com/?p=8#comment-2 [...] at 7:34 pm · Filed under Blogs, Statistics ·Tagged comment counts Last week, I wrote a very technical post on how I model the distribution of comment counts for an RSS feed in FeedHub.  I originally [...] [...] at 7:34 pm · Filed under Blogs, Statistics ·Tagged comment counts Last week, I wrote a very technical post on how I model the distribution of comment counts for an RSS feed in FeedHub.  I originally [...]

]]>