Saturday, September 6, 2008

Gender Guesser

Note: This looks familiar, so I've probably seen it before. Whether I've gotten around to writing about it is another matter. It's unlikely to interest anyone but me, so don't hesitate to skip it. :o)

I happened upon a "Gender Guesser", recently. No, it's not based on appearance. (If it were, I'd be mightily offended. Read on to see why. (g)) I won't try to explain how it works (because I don't really know), but basically you plug a chunk of writing into this program, and it tries to guess whether the author is/was male or female. There's one "score" for formal writing and another for informal writing. It's an interesting concept, so of course I had to give it a try. (And then another. And another.)

First, I tried my last entry (barring the part where I briefly quoted someone else). Results for informal: weak female. (No, that doesn't mean that I lack strength or moral integrity. ;o) It indicates that, according to the program, my writing style registered as female, but not emphatically so.) The "score" was 1009 female vs. 828 male, to be precise.

(It strikes me as odd that the "formal" score for that same text scored as weak male, meaning if I'd written it for something other than a casual blog-- as an essay, for example-- they'd guess it was written by a male. I wonder why would that be. . .)

Next, I went one entry further back, to the story about the trials and tribulations of life with Trixie. Results for informal: 607 f to 1186 m, male. Formal: 744 f to 717 m, weak female.

I skipped a couple of entries, then, because they were under the recommended 300-word minimum. So next up was a bulleted list. Very informal. Informal score: 336 f vs. 644 m, male. (The formal score was weak male.)

(By now, I was obsessed, so I kept going, even though the notes on the Gender Guesser page indicate that numerous texts written by the same person are likely to yield similar results. (g))

Entry about Gustav, Trixie and car tags:
Informal: 1120 f vs 1573 m, weak male
Formal: 941 f vs 846 m, weak female

"Disconnected Thoughts" entry:
Informal: 855 f vs 1271 m, weak male
Formal: 693 f vs 653 m, weak female

"Pupdate" entry:
Informal: 589 f vs. 588 m, weak female
Formal: 777 f vs. 416 m, female

"Random / what else is new" entry:
Informal: 3078 f vs 3606 m, weak male
Formal: 1947 f vs. 2230 m, weak male

So. . . Most of my "informal" results are male (albeit a weak male ;o)), while the "formal" results tend more toward weak female. Either my "informal" writing style should be classified as "formal" or I have a somewhat masculine style of writing.

Then there's the whole European English vs. American English element: "In general, if the difference between male and female weight values is not significant (a "weak" score), then the author could be European." Of course, obviously I'm not European, so. . . (shrug)

There's also this: "Professional female writers (and experienced hobbyists) frequently use male writing styles." I'm by no means a professional writer, but maybe I qualify as an experienced hobbyist.

Anyway, however unreliable it may be, the program served as a nice little amusement. :o)