Using your browser URL history to estimate gender | Hacker News Reader

Using your browser URL history to estimate gender

mikeonads.com

37 pointspaulgb22 comments

Probably safe enough for them to always say "male." The hit rate would be pretty high given the type of person hitting that page...

They show how they got the result.

wouldn't this be easy?

1) Search browser history, look for porn. 2) If porn site = true, gender = male 3) else gender = female

Not really because usually that behavior is followed by "Clear History" in a browser your wife does not know is installed but may stumble upon some day. :)

(Or you work at a corporate job.)

I don't visit porn sites and I'm male.

I never got any of my porn through websites.... I mean... uh, what's porn?

you date prudes.

Likelihood of you being FEMALE is 0% Likelihood of you being MALE is 100%

Oh.

Apparently I'm 59% female. Hmm. I could have sworn...

Has anyone considered that this is a massive privacy issue?

Imagine an advertiser detecting which sites you've been to, and serving different ads based on what it finds.

Edit: the author of the script thought of it: http://azarask.in/blog/post/socialhistoryjs/ and it's bugzilla 147777 https://bugzilla.mozilla.org/show_bug.cgi?id=147777

Perhaps I'm in the minority here, but I have far larger privacy concerns than a website being able to gestimate my gender.

Yeah, it's a pretty massive privacy issue, but at least I would hope people wouldn't do it and expect to get away with it.

I would be pretty surprised if people weren't already doing this.

It would be interesting to find out. If any of the big players are doing it, there would be a huge backlash if they got caught for it.

Ad companies already use cookies all over the place to track users, I'm not sure this would be that much bigger of a deal.

Likelihood of you being FEMALE is 99% Likelihood of you being MALE is 1%

You might want to try making contact with other users of the sites you frequent... ;)

Hmm... From the comments, it seemed to favor female...

Not a problem for me, though, since newegg.com, pricewatch.com, ubuntuforums, etc all have 2+ ratios of men to women. Hah. (I'm male)

Resulst seem pretty suspect:

Likelihood of you being FEMALE is 57%

Likelihood of you being MALE is 43%

Site Male-Female Ratio

google.com 0.98

commerceonline.com 0.77

I'm female, but got these results -

Likelihood of you being FEMALE is 45% Likelihood of you being MALE is 55%

Apparently, my surfing habits are androgynous.

Another reason to use NoScript.

Or Opera.

Reply on news.ycombinator.com