There’s an interesting post and Java applet over on Mike’s site that examines the idea that you can determine a person’s gender by examining their browser history.

” I… apply the ratio of male to female users for each site and with some basic math determine a guestimate of your gender. The math is really quite simple, I just take:

1 / (1 + r_1 * r_2 * … * r_n)
where p_i is the ratio of men-to-women for the specific site. For example, if you had been to two sites that had a 2-1 ratio of men to women, the probability of you being female would be:
1 / (1 + 2 * 2) = 1/5 = 20%”

That doesn’t mean anything at all to me, because I suck at maths, but the Java application is quite interesting. There’s a button on the site that lets you perform an analysis on your own browser. Mine was accurate.

Check it out here. Sent in by Abby - Thanks Abs!

  • E-mail this story to a friend!
  • Digg
  • StumbleUpon
  • del.icio.us
  • Reddit
  • Facebook
  • Mixx
  • Blogsvine
  • Fleck
  • Live
  • laaik.it
  • Y!GG
  • NewsVine
  • Technorati