?

Log in

No account? Create an account

Previous Entry | Next Entry

weather: sunny/cloudy
outside: 14°C
mood: amused
Bookblog's Gender Genie algorithm, as implemented by hutta, pegs me as male. Ha, you'd think with a name like "bride", there would be no mistake =) But from the looks of it, it's fairly inaccurate, especially with women. You'll also notice that statistically, the number it got right vs. wrong (35%/58%) looks almost the same as the male/female split (36%/63%) from the LJ Statistics page. This means that his tool could just be saying that everyone is male and it's correct only insofar as it's mirroring the population. =)

I give hutta's coding ability the benefit of the doubt here. Let's say he didn't screw up something major, like put a '>' where it was supposed to be a '<' or vice versa (which it almost looks like he has; if he'd just reverse the last comparison operator in his code and call everyone "female", then he'd be statistically more accurate than not) =)

Gender Genie's interpretation of the Koppel-Argamon paper is also an unknown. It does say that they use "a simplified version of an algorithm developed by Moshe Koppel, Bar-Ilan University in Israel, and Shlomo Argamon, Illinois Institute of Technology..."

There's also the soundness of the paper itself that's in question. However widely accepted research is, the conclusions drawn from their test cases can be too general, too specific or dead wrong. Natural Language Processing is a veryvery difficult thing to get right. I haven't looked at the paper yet, I just printed it out, so my ass and my elbow could be looking a bit similar right now =)

I've found that informal writing is also very prone to throwing off these kinds of algorithms. For example, in a blog, both men and women usually tend to abbreviate, especially with numbers. Well, if you're thinking that men use numbers and women spell them out, everyone will look male to you. As well, blogs and journals usually have datestamps and timestamps on them, which are usually numbers. Gender Genie claims to have a different algorithm that takes care of this case, but they're still trying to predict humans, so I still have my doubts.

Nonetheless, props to hutta for the implementation of an intriguing concept =)

Tags:


Comments

( 15 comments — Leave a comment )
magicwoman
Sep. 19th, 2003 11:05 am (UTC)
Good afternoon!
Don't feel bad, I turned out male too. Go figure, a girly girl like me?

Susan
bride
Sep. 19th, 2003 12:20 pm (UTC)
Re: Good afternoon!
Exactly. It's saying EVERYONE'S male =) So, statistically, the number it got right (35%/58%) looks almost the same as the male/female split (36%/63%) from the LJ Statistics page.
serennig
Sep. 19th, 2003 11:19 am (UTC)
I don't place much stock in anything that tries to tell girls from boys... well, 99% of the time I will trust a pelvic check, but that's not very convenient in most situations. But beyond the realm of reproduction, I don't figure there's any real difference. Who a person is is such a complex thing; focussing on any one part -- gender, race, culture, place of birth, education, parenting, age -- makes one miss out on the other five trillion parts.

Though it does make for an amusing meme. :P
bride
Sep. 19th, 2003 12:21 pm (UTC)
Exactly. It's saying EVERYONE'S male =) So, statistically, the number it got right (35%/58%) looks almost the same as the male/female split (36%/63%) from the LJ Statistics page.

Pelvic check... haha =D
serennig
Sep. 19th, 2003 12:38 pm (UTC)
Do you mean, the author rigged it to not actually do any algorithming, but just randomly say "you're xx% male"? Or just that it's saying a lot of people are male? *confused*

If it's the latter, I'd put money on the over-weighting of the word "the" (supposedly a masculine word) for causing the skewing. If it were truly and accurately based on retrospective analyses of male/female writings, you'd think we'd see a more equal distribution.

Unless LJ users are just skewed to the supposedly masculine side for some reason.
bride
Sep. 19th, 2003 12:45 pm (UTC)
Do you mean, the author rigged it to not actually do any algorithming, but just randomly say "you're xx% male"?

I'm joking that the evidence seems to point to exactly that, yes =)

Unless LJ users are just skewed to the supposedly masculine side for some reason.

That's another interesting thought because most LJers are female. All of our journal entries have characteristics that are traditionally considered male?
(Deleted comment)
bride
Sep. 19th, 2003 12:23 pm (UTC)
I haven't yet met anyone it picked as female, whether they were or not.

Exactly. It's saying EVERYONE'S male =) So, statistically, the number it got right vs. wrong (35%/58%) looks almost the same as the male/female split (36%/63%) from the LJ Statistics page.

*fist in the air* GIIIRL!!! =)
astral
Sep. 19th, 2003 01:04 pm (UTC)
lol. it says i'm female on the journal i write in, and quite a few of my female friends too.
bride
Sep. 19th, 2003 01:18 pm (UTC)
Haha =) Well, there goes that theory... =)
astral
Sep. 19th, 2003 09:31 pm (UTC)
did u think i was male??
bride
Sep. 19th, 2003 09:33 pm (UTC)
Nono =) I knew you were female, but it puts a wrench in my theory that the quiz says _everybody_ is male =)
(Deleted comment)
bride
Sep. 19th, 2003 03:14 pm (UTC)
Yeah, only one person on my friends list came up female.

And I'll bet it was a guy... =D
(Deleted comment)
bride
Sep. 19th, 2003 03:46 pm (UTC)
Whoa... =O
bdspitapit31
Sep. 20th, 2003 11:28 am (UTC)
Wow. It actually said I was a girl. I was fully prepared to be a member of the male species, too.
bride
Sep. 20th, 2003 05:13 pm (UTC)
Hmm... bizarre...
( 15 comments — Leave a comment )

Profile

eLouai
bride
The Bride of the First House

Latest Month

March 2015
S M T W T F S
1234567
891011121314
15161718192021
22232425262728
293031