Using your browser URL history to estimate gender
July 13th, 2008
Thanks to Paul Cook for the initial link to this fascinating little javascript script Social History. Thes cript analyzes the css color of various links to determine whether or not the user has been to that site. If the link has the “visited” style, then he marks the user as having been to that site. Now the Social History implementation of this is rather innocuous — it’s a clever way of only displaying only the sharing buttons of sites that the user is an active participant of. Of course there are far more interesting applications for advertising.
One of the things that I always wanted to do but never got around to was to analyze a user’s browsing history to estimate age and gender. Of course the idea is definitely not new, in fact Xerox (of all companies??) has a patent on the whole process and I’m certain plenty of networks already do something of the sort… but what the heck, let’s have some fun!
So what I did is I modified the SocialHistory JS so that it polled the browser to find out which of the Quantcast top 10k sites were visited. I then 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:
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:
Ok, so Click the button to give it a shot (those of you using RSS readers probably need to click this link to open this post in a browser):
UPDATE: This takes a while on Internet Explorer — please be patient (or try FireFox)
Kind of cute right? Don’t worry — I am not storing your history in any way, this is purely for fun. I’d appreciate it if you paste the resulting probabilities in the comments together with your actual gender, I’m interested to know the accuracy of this simplistic approach. In case it isn’t obvious — please don’t do this for real.
Related Posts:
- A Tag Cloud using the del.ico.us API + my Gender Hack
- Is Google taking behavioral data to display?
- How do behavioral networks work?
- It’s time to drive for simpler integration
- Microsoft is only 6 years late on behavior
July 13th, 2008 at 7:08 pm
Mike, it is a product of r1*r2*… and not a sum.
July 13th, 2008 at 7:14 pm
Thanks, I think I just found out something about myself. Apparently your little test thinks I am 91% female in my browsing behavior.
July 13th, 2008 at 7:31 pm
Thanks for catching the typo — fixed. Note the actual calculation was correct.
July 13th, 2008 at 8:49 pm
Hmmm, your test thinks I am:
Likelihood of you being FEMALE is 99%
Likelihood of you being MALE is 1%
July 13th, 2008 at 9:14 pm
Btw, I never did post my own stats!
Likelihood of you being FEMALE is 26%
Likelihood of you being MALE is 74%
July 13th, 2008 at 9:24 pm
75% male.
Its interesting. I never realized the major websites actually had a female skew.
Yahoo .9
Amazon .9
Pandora .9
Gmail .9
July 14th, 2008 at 12:39 am
Says I’m male (59% to 41%). Guess that puts those lingering gender identity issues to bed…
July 14th, 2008 at 12:44 am
Likelihood of you being FEMALE is 62%
Likelihood of you being MALE is 38%
Lucky for your test I surfed Ikea and Target this wknd instead of my true internet love ESPN.
July 14th, 2008 at 1:39 am
Hmph…
Likelihood of you being FEMALE is 17%
Likelihood of you being MALE is 83%
July 14th, 2008 at 3:33 am
[...] Mike On Ads: Use your web browser history to estimate whether you’re male or female. It looks at the sites you’ve visited, uses the typical demographics for those sites, plus a [...]
July 14th, 2008 at 6:01 am
Nice post Mike. I got mine right. Cool stuff. Love it.
Thanks,
Mukul.
http://mukulblog.blogspot.com
July 14th, 2008 at 6:41 am
Results: FEMALE is 44%/MALE is 56%
July 14th, 2008 at 7:53 am
Likelihood of you being FEMALE is 3%
Likelihood of you being MALE is 97%
July 14th, 2008 at 8:47 am
Likelihood of you being FEMALE is 87%
Likelihood of you being MALE is 13%
July 14th, 2008 at 8:52 am
Another important thing to note — if I could only grab the FREQUENCY of visits to a URL the estimates would be a lot more accurate. Of course all you need to do that is a tag present on a lot of these sites…
*hmmmm*
July 14th, 2008 at 9:05 am
My gender identify issues remain unresolved:
Likelihood of you being FEMALE is 50%
Likelihood of you being MALE is 50%
Guess I’m just a person. The Stephen Colbert of gender.
July 14th, 2008 at 3:56 pm
100% male, I guess there is no doubt
July 14th, 2008 at 5:25 pm
Likelihood of you being FEMALE is 54%
Likelihood of you being MALE is 46%
I’m male.
July 14th, 2008 at 6:03 pm
Likelihood of being Female is 0% and of being male is 100%. I’m Male.
July 14th, 2008 at 7:03 pm
Likelihood of you being FEMALE is 1%
Likelihood of you being MALE is 99%
I’m a proud male!
It works perfectly! haha!
July 14th, 2008 at 7:04 pm
Likelihood of you being FEMALE is 44%
Likelihood of you being MALE is 56%
(I’m female).
July 14th, 2008 at 7:19 pm
Likelihood of you being FEMALE is 75%
Likelihood of you being MALE is 25%
Let me check… (looks in pants), yep, still a make.
July 14th, 2008 at 7:20 pm
Likelihood of you being FEMALE is 2%
Likelihood of you being MALE is 98%
I’m a guy - very impressive.
July 14th, 2008 at 7:44 pm
Likelihood of you being FEMALE is 2%
Likelihood of you being MALE is 98%
and i’m male.
July 14th, 2008 at 7:50 pm
Likelihood of you being FEMALE is 5%
Likelihood of you being MALE is 95%
I’m female.
July 14th, 2008 at 7:50 pm
100% male.. and I am
July 14th, 2008 at 7:55 pm
96% Male and you guessed it, I’m a guy.
July 14th, 2008 at 7:59 pm
YES!!!!! WOOOHOOOO!! I’m so pumped right now.. I’ma gonna to mack down some power bars and go shoot some deer…
Likelihood of you being FEMALE is 0%
Likelihood of you being MALE is 100%
July 14th, 2008 at 8:00 pm
Likelihood of you being FEMALE is 28%
Likelihood of you being MALE is 72%
Male.
July 14th, 2008 at 8:05 pm
Likelihood of you being MALE is 100%
100% correct!
July 14th, 2008 at 8:09 pm
i feel too close for comfort - male:
Likelihood of you being FEMALE is 44%
Likelihood of you being MALE is 56%
July 14th, 2008 at 8:27 pm
It said there was a 87% chance I’m female, and I am. But I admit that women in San Francisco often seem attracted to me. That probably explains the other 13%.
Very cool.
July 14th, 2008 at 8:27 pm
I think there’s something wrong with your widget. Or else I’m hanging out in the wrong places altogether.
I’m male. My wife would be greatly surprised by these results:
Likelihood of you being FEMALE is 75%
Likelihood of you being MALE is 25%
July 14th, 2008 at 8:37 pm
Wow
Likelihood of you being FEMALE is 100%
Likelihood of you being MALE is 0%
I’m female
July 14th, 2008 at 8:41 pm
Likelihood of you being FEMALE is 22%
Likelihood of you being MALE is 78%
July 14th, 2008 at 8:43 pm
Likelihood of you being FEMALE is 24%
Likelihood of you being MALE is 76%
No, it’s quite sure that I’m female. I had a doctor check and everything.
July 14th, 2008 at 8:50 pm
Likelihood of you being FEMALE is 37%
Likelihood of you being MALE is 63%
I’m male but I need to check my browsing history!
July 14th, 2008 at 8:53 pm
i am a 36 year old male:
Likelihood of you being FEMALE is 67%
Likelihood of you being MALE is 33%
July 14th, 2008 at 8:55 pm
Missed it for me. Sorry, gender stereotypes.
Likelihood of you being FEMALE is 37%
Likelihood of you being MALE is 63%
July 14th, 2008 at 9:17 pm
Likelihood of you being FEMALE is 57%
Likelihood of you being MALE is 43%
Site Male-Female Ratio
yellowpages.com
0.72
wachovia.com
0.87
megaupload.com
1.5
pnc.com
0.89
gamefly.com
0.83
nexon.net
1.08
The correct answer…
*dramatic pause*
Male!
July 14th, 2008 at 9:18 pm
97% Male. Correct.
Interesting the Gizmodo is more male than Engadget.
July 14th, 2008 at 9:34 pm
Likelihood of you being FEMALE is 51%
Likelihood of you being MALE is 49%
By a slight margin, it gets it right. I’m intrigued that it’s that narrow a window.
July 14th, 2008 at 9:37 pm
Amazingly, it also guessed I’m a software engineer, as it wants me to do some debugging….
Warning: Invalid argument supplied for foreach() in /homepages/45/d144749141/htdocs/moa/gender/analyze.php on line 23
Likelihood of you being FEMALE is 50%
Likelihood of you being MALE is 50%
Site Male-Female Ratio
July 14th, 2008 at 9:38 pm
I’m male but:
Likelihood of you being FEMALE is 55%
Likelihood of you being MALE is 45%
If I surf a few porn sites and come back will it put me over the top?
July 14th, 2008 at 9:44 pm
Likelihood of you being FEMALE is 2%
Likelihood of you being MALE is 98%
Thanks Jegs and Summit Racing!
July 14th, 2008 at 9:46 pm
I am all that is man!!! (Yes I’m Male)
Likelihood of you being FEMALE is 0%
Likelihood of you being MALE is 100%
Site Male-Female Ratio
google.com 0.98
yahoo.com 0.9
myspace.com 0.74
youtube.com 1
ebay.com 1.11
amazon.com 0.9
walmart.com 0.77
bestbuy.com 1.11
chase.com 0.87
circuitcity.com 1.2
wunderground.com 1.22
linkedin.com 0.94
gizmodo.com 2.08
macrumors.com 2.08
slickdeals.net 1.27
gametrailers.com 1.53
godaddy.com 1.17
fatwallet.com 1.38
footlocker.com 0.72
jalopnik.com 2.13
ufc.com 1.35
engadget.com 1.7
tinyurl.com 0.83
compusa.com 1.7
joystiq.com 1.44
sherdog.com 2.33
logmein.com 1.17
bmwusa.com 1.41
July 14th, 2008 at 9:59 pm
Lonn doesn’t give away the sex right away, but it’s male. Time to go hit ESPN and manly sites to get my numbers straight
Likelihood of you being FEMALE is 62%
Likelihood of you being MALE is 38%
Site Male-Female Ratio
google.com
0.98
yahoo.com
0.9
facebook.com
0.83
tinyurl.com
0.83
July 14th, 2008 at 10:03 pm
I’m male but my history says I’m not. now I’m confused, lol
Likelihood of you being FEMALE is 97%
Likelihood of you being MALE is 3%
Site Male-Female Ratio
google.com
0.98
youtube.com
1
ebay.com
1.11
facebook.com
0.83
walmart.com
0.77
blogger.com
1.06
target.com
0.67
flickr.com
1.15
apple.com
0.89
careerbuilder.com
0.79
lowes.com
0.98
wunderground.com
1.22
foodnetwork.com
0.63
kohls.com
0.57
directv.com
0.87
wachovia.com
0.87
veoh.com
1.06
mac.com
0.9
linkedin.com
0.94
samsclub.com
0.96
tmz.com
0.77
newgrounds.com
0.85
sportsline.com
1.86
salliemae.com
0.63
mycokerewards.com
0.79
nwa.com
0.82
bellaonline.com
0.64
emusic.com
0.85
smilebox.com
0.57
spock.com
1.15
blogcatalog.com
1.13
audible.com
0.82
naymz.com
1.06
vimeo.com
1.15
ustream.tv
1.33
stickam.com
1.15
July 14th, 2008 at 10:05 pm
I’m female, but according to your estimate:
Likelihood of you being FEMALE is 50%
Likelihood of you being MALE is 50%