Tuesday, August 03, 2010

CS: The popularity of blog entries and the Big Picture

On June 30, Google decided to provide specific statistics on the hits generated on specific blog posts. By that I mean that while a number of you get your information from Feedreaders and Email, a certain number of readers come directly to the site either because they are new or because they want to read the post further. I knew about some of these numbers for some early posts, but I was still surprised at how large they were compared to the everyday flux of visitors (Data gathering was performed from June 30 till July 15th).

In the top ten, we have two entries that were written about three years ago. One features videos made at the Summer school on Compressive Sampling and Frontiers in Signal Processing that took place at IMA on June 4-15, 2007 , the other one features an example of compressed sensing (with code) that doesn't rely on Gaussian matrices. This last entry garners also the most comments centered around the inability to install SeDuMi correctly. Maybe the example should be changed and rely only on scripts that can run directly with Matlab or Octave ? Of note, these two entries are part of the Big Picture introduction which itself is linked from Wikipedia.

Remember this cartoon ?












Well, the stats show a jump after it was also featured on the Big Picture as well. So while the Wired article brought some new readers, the jump in readership drew down pretty quickly whereas the wikipedia entry provides a seemingly constant channel of interested readers. If you are interested in seeing the number of people coming on wikipedia page on compressed sensing, you may want to go there. One last thing, I also set up an Adsense account and you may have noticed ads on blog entries and entries showing up in your feeds. I reasoned that since money was involved the pageviews and attendant statistics would be providing a serious and instantaneous picture of the interest in the blog and attendant pages. Here is the stat of a typical weekday:


and the same stat about a week later:



Or about 1700 pageviews per day. It doesn't translate into 1700 readers per day since many of  the readers navigate between the blog entries and the Big Picture. I currently estimate the traffic to be about 1200 readers per day.

No comments:

Printfriendly