Male or Female? Your Browser’s History Might Tell

A website should not be able to read the the history of websites that a user visited previously. An ingenious CSS hack and some spicy Javascript is however enough to use a trial and error method to find out if a user has been to specific websites.

Some webmasters use this method to display the social bookmarking and news batches of websites that their visitors have been through which is very effective in gaining popularity on those websites.

Others came up with a script that is using this method to estimate if the visitor is male or female. The Javascript makes use of the Quantcast Top 10K websites and the male:female ratio that they provide plus a neat algorithm to compute the percentages for male and female likeliness.

Mine turned out 50/50 because I don’t use the history at all. Good luck and let me know what the script thinks about your gender..

The same principle could possible be used to display targeted advertisements. Have been to a lot of parenting websites lately? How about parenting ads? Like sports? How about an ESPN subscription? It’s a bit scary, don’t you think?

Related Articles:

Author: Martin Brinkmann, Monday July 28, 2008 -
Tags:code, gender, history, javascript, quantcast

Responses so far:

1. kazounet says:

   July 28, 2008 at 4:54 pm

   50/50 for me too, though I have a concequent firefox history :/

   Reply
2. Jonathan says:

   July 28, 2008 at 5:22 pm

   Likelihood of you being FEMALE is 54%
   Likelihood of you being MALE is 46%
   Odd… :)

   Reply
3. Thinker says:

   July 28, 2008 at 6:59 pm

   Likelihood of you being FEMALE is 2%
   Likelihood of you being MALE is 98%
   Hehe :D Mine top male sites:
   Site Male-Female Ratio
   thepiratebay.org 2.13
   leftlanenews.com 2.23 ^^

   Reply
4. Rarst says:

   July 28, 2008 at 9:09 pm

   96% Male

   Too much software portals… :) And the method is seriously cool and clever.

   Reply
5. JoeSurfer says:

   July 30, 2008 at 1:18 am

   I had 100% of being female… news to me seeing as I’m a guy. The method might be cool but perhaps there’s some assumptions made on how ‘male’ or ‘female’ a site is?

   Reply