Comment Analysis

So I was reading my gMail just the other day (what's new about that? It seems I'm always ready email) and I was distracted by a WebClip (the updates feed by a few selected feeds at the top of my mail plan) about people angry over a Chaser skit I saw the night before. I clicked thought and read about how people had got their feathers ruffled over an "anti-ad" against the Make a Wish Foundation. I wasn't surprised, this is something the ABC and Chaser crew must be used to by now. I'm actually starting to think that the organisations complain deliberately and publicly just to get some free PR out of it.

I noticed that the article was only written just after an hour ago. I read the article and then noticed that the comments were already closed. I thought "that's strange". Most comment systems close after a few week or months to stop old debates being rehashed as after a certain period of time it become irrelevant.

But within a few hours is abnormal. So I had a look at the number of responses and there was an amazing 225 comments already. This must be the ABC limit of comments on a single article - and fair enough - looking at most of the posts it was easy to see it just became a case of tit-for-tat and opinions were just being repeated (nothing new, nothing interesting).

But what really stuck out to me was the number of comments in a very short time period. In about 80 minutes there had been 225 comments. That mean that there was nearly 3 comments submitted every minute. And I didn't that that the ABC website had that many users (well not in comparison to Ninemsn for news).

It then got me thinking… I wonder what a graph of comments over time would look like. Did it start slow and then build until the cut off? or were people very vocal as soon as the news "went to press" and then did it die off as similar opinions were expressed? I wonder?

So I thought I should be able to find out quite easily and then also compare this to different types of sites. Were the projectiles of comment frequency similar on other sites? How would the ABC compare to the BBC? Or Engadget? Or shouts on


From using jQuery I know it's a simple task to craw the DOM that makes up a page a collect the data that I needed. I also know that Google provides an easy to use Charts API that would allow me to visually plot the data. Then I needed as easy way to run the code on the 3rd party sites. I didn't want the lock down to one browser (vendor or version) so I choose to make a favelet (or Bookmarklet - call it what you may)

And here it is.

Behind the News

Here's how I built it.

R Walker
analytics, blog, comments, data, favlet, javascript, jquery

Why does this site look like this? Your browser has become outdated and does not support recent web standards. You're welcome to use this page as is, although we suggest upgrading your browser. This page is designed to look it's best in Firefox although it will work as intended in the latest version of most popular browsers.