LOL. This doesn't surprise me at all. I totally fit the bill. I use emotes and abbreviations ALL the time. Maybe I should stop just so I don't perpetuate gender stereotypes
Don't worry, there are plenty of middle-aged white men in Ohio who will perpetuate stereotypes for you.;)
The headline is crap. Anyone who would use the algorithm in that manner is likely to be wrong at least once if they used it more than twice. That's not what it does nor what it's meant to do. It's like browsing habits -- making guesses about the browser based on where they visit is good enough to improve advertising effectiveness. It does not definitely tell you anything at all about any specific individual, ever.
The reporting on this is awful, but I don't see how the study itself is poorly constructed. Your objection is largely irrelevant. "Basically it assumes that anyone attempting to impersonate the opposing gender is a tech ignorant moron that has made no effort to create a persona - something that is contrary to pretty much every piece of information we have on people who do this." No. Basically, it assumes the number of people doing this are not a large enough percentage of the data to have any significant impact on the results. A statistical study is not invalid because there's a bit of noise in the data. Indeed, practically no statistical study would be possible if that were the case. It makes no assumptions at all about people attempting to impersonate the opposing gender beyond that they're rarer than people who aren't. As long as that fact is true, the methodology used here is sound, although the more people who are, the worse the results should be unless they're doing it perfectly.
Like most statistical things like this, it will go off-the-rails when fed input from minorities or other cultures. That doesn't invalidate it, as it's largely irrelevant for its intended uses (e.g. ad targeting -- no one cares if it's not 100% accurate, it only needs to work for the majority and be accurate enough to increase revenue more than the cost).
If you're like me, your whole behavior probably shifts significantly depending on context as well, so anything like this would likely render different judgements depending on when used.
You do, or you wouldn't have clicked into this article to read and comment on it. By the way, by doing so, you've indicated the site should post more content like it. Simply clicking is a metric sites use for determining the interest in stories, but commenting is an even bigger and more meaningful metric. What precisely you say is irrelevant, the fact that the story generated more discussion is all the number-cruncher sees.
Or it can be used as a training tool for would-be impersonators.
It probably doesn't tell them anything they haven't already gathered simply by watching people communicate. I don't think a single rule mentioned was anything I haven't noticed myself. Anyone who would have a good chance of pulling it off already knows this stuff and more.
What about metrosexual or gay men/woman as they cross the lines.
The buying habits of these people is probably as skewed from gender-based stereotypes as their communication styles. If their ad experience ends up being skewed by using this algorithm, that's probably a feature, not a bug.
Anyone trying to define sexuality as a yes or no (just ask a biologist what is natural and they will laugh at you) based on a set language subset is in for a demoralising hiding.
Irrelevant, not what they're trying to do, and I suspect you've missed the point entirely...
How much of this is universal and just American-English? Can this be applied to other languages to pick up like patterns?
This in particular would only work on American English-speakers or sufficiently similar, but the same idea should work anywhere when modified for the local culture, and in many cultures, probably be more accurate, as communication patterns follow who you interact with regularly, and in many cultures, cross-gender communication is less common, more limited, and/or significantly less egalitarian.
Presumably the people tweating like a girl want their targeted ads to skew that way anyhow. Heck, they're probably more likely to buy that way, too...
(Sorry, were you thinking this research has any use or point outside automated online ad delivery systems? It has no possible application that involves making any kind of judgement about any specific individual.)
Well, yes, obviously. It's useful for ad-targeting, which needs to be automated. Using it in any context where it's feasible to just have a human read the text is pointless, and there's obviously no way to conclusively determine anyone's gender without requiring them to disrobe. Far too many men chat like women, and women chat like men. Like virtually everything even remotely like this, it's useful when applied to groups for a statistical measure, and utterly useless when applied to any one individual to determine anything about them. Not everyone is the epitome of their stereotype.
No. We hypothesized this already. We still don't know it, but we now have better reason for believing it than "it sounds right". Silly scientists are more interested in truth than truthiness.
Mind you, Im not sayiing they private companies shouldnt exist. Unlike the prisons and military, there is a place for private spaceflight orgs providing other private organizations with launch vehicles. But government....no way.
I'm unclear why it's okay for NASA to pay Boeing/Rockwell for an orbiter and Thiokol/Allianet for boosters (aka how they've been operating the manned space program since the 80s), but if they go with one company (SpaceX) for an entire system instead, this is suddenly a bad idea?
The problem with studies like this is, people look at the results and conclude, "So, if everyone used their real name in the Internet, it wouldn't be a problem." Try again. The same deindividualization would still occur. John Andersen is just as effectively anonymous whether he posts as John Andersen or TrollFace23. Uniquia McRarename may have a problem, but the remaining 98% of the forum users won't.
What I find most interesting is that nobody has mentioned Diaspora since Google+ launched.
That's because nobody mentioned Diaspora before Google+ was launched, either.
In response to your comment, I have now mentioned it, but I honestly have no idea what we're talking about. I know what "diaspora" means, of course, but the only time I've ever seen it used with a big-D is when referring to the Jewish diaspora following the Babylonian exile. From the context in which you're using it, I'm guessing it's the name of some other social networking product that most people have never even heard of?
That means it's sure to win. This reminds of of when OS/2 mopped the floor with Windows because it had superior multitasking and memory management!
If only they'd had a superior product... or at least a decent UI. Lacking that, all the wonderful multitasking and memory management in the world was pretty useless.
Slashdot Mirror
LOL. This doesn't surprise me at all. I totally fit the bill. I use emotes and abbreviations ALL the time. Maybe I should stop just so I don't perpetuate gender stereotypes
Don't worry, there are plenty of middle-aged white men in Ohio who will perpetuate stereotypes for you. ;)
The headline is crap. Anyone who would use the algorithm in that manner is likely to be wrong at least once if they used it more than twice. That's not what it does nor what it's meant to do. It's like browsing habits -- making guesses about the browser based on where they visit is good enough to improve advertising effectiveness. It does not definitely tell you anything at all about any specific individual, ever.
Indeed. Only an idiot sees targeted ads as more evil than untargeted ads.
The reporting on this is awful, but I don't see how the study itself is poorly constructed. Your objection is largely irrelevant. "Basically it assumes that anyone attempting to impersonate the opposing gender is a tech ignorant moron that has made no effort to create a persona - something that is contrary to pretty much every piece of information we have on people who do this." No. Basically, it assumes the number of people doing this are not a large enough percentage of the data to have any significant impact on the results. A statistical study is not invalid because there's a bit of noise in the data. Indeed, practically no statistical study would be possible if that were the case. It makes no assumptions at all about people attempting to impersonate the opposing gender beyond that they're rarer than people who aren't. As long as that fact is true, the methodology used here is sound, although the more people who are, the worse the results should be unless they're doing it perfectly.
Linguists know that "out" is not a verb.
Only historical ones, or very bad ones. Modern ones who are good at their job know that it is.
Like most statistical things like this, it will go off-the-rails when fed input from minorities or other cultures. That doesn't invalidate it, as it's largely irrelevant for its intended uses (e.g. ad targeting -- no one cares if it's not 100% accurate, it only needs to work for the majority and be accurate enough to increase revenue more than the cost).
If you're like me, your whole behavior probably shifts significantly depending on context as well, so anything like this would likely render different judgements depending on when used.
And here I thought I had solidified my masculinity when I burnt all my pink shirts.
At least you didn't burn your bras... ;)
Really. Who cares?
You do, or you wouldn't have clicked into this article to read and comment on it. By the way, by doing so, you've indicated the site should post more content like it. Simply clicking is a metric sites use for determining the interest in stories, but commenting is an even bigger and more meaningful metric. What precisely you say is irrelevant, the fact that the story generated more discussion is all the number-cruncher sees.
Or it can be used as a training tool for would-be impersonators.
It probably doesn't tell them anything they haven't already gathered simply by watching people communicate. I don't think a single rule mentioned was anything I haven't noticed myself. Anyone who would have a good chance of pulling it off already knows this stuff and more.
What about metrosexual or gay men/woman as they cross the lines.
The buying habits of these people is probably as skewed from gender-based stereotypes as their communication styles. If their ad experience ends up being skewed by using this algorithm, that's probably a feature, not a bug.
Anyone trying to define sexuality as a yes or no (just ask a biologist what is natural and they will laugh at you) based on a set language subset is in for a demoralising hiding.
Irrelevant, not what they're trying to do, and I suspect you've missed the point entirely...
It seems to me that...
I love it when people attempt to make statistical generalizations from personal experience. XD
How much of this is universal and just American-English? Can this be applied to other languages to pick up like patterns?
This in particular would only work on American English-speakers or sufficiently similar, but the same idea should work anywhere when modified for the local culture, and in many cultures, probably be more accurate, as communication patterns follow who you interact with regularly, and in many cultures, cross-gender communication is less common, more limited, and/or significantly less egalitarian.
Presumably the people tweating like a girl want their targeted ads to skew that way anyhow. Heck, they're probably more likely to buy that way, too...
(Sorry, were you thinking this research has any use or point outside automated online ad delivery systems? It has no possible application that involves making any kind of judgement about any specific individual.)
How is this "huge?" What the hell are you going to do with it? Someone tweets and uses an exclamation point, so you... what now?
Display the ad for the family car instead of the sports car. Duh...
Well, yes, obviously. It's useful for ad-targeting, which needs to be automated. Using it in any context where it's feasible to just have a human read the text is pointless, and there's obviously no way to conclusively determine anyone's gender without requiring them to disrobe. Far too many men chat like women, and women chat like men. Like virtually everything even remotely like this, it's useful when applied to groups for a statistical measure, and utterly useless when applied to any one individual to determine anything about them. Not everyone is the epitome of their stereotype.
Unless it's in the context of things like "There's no way I'm about to be eaten by a slavering xenomorph!"
Okay... shouldn't your name be Carter Burke? ;)
Not impossible, it just requires getting to the other side of the valley. This just shows that, for these purposes, a 90% solution isn't acceptable.
The feeling of being creeped-out by a NON-moving humanoid.
Actually, it does. A non-moving humanoid definitely violates our notions of what kind of motion we expect from a human.
We knew this already.
No. We hypothesized this already. We still don't know it, but we now have better reason for believing it than "it sounds right". Silly scientists are more interested in truth than truthiness.
They spent money to scientifically determine 'it looks weird'.
Actually, no. If it "looked weird", that wouldn't be a problem. The problem occurs when it looks normal, but moves weird.
Mind you, Im not sayiing they private companies shouldnt exist. Unlike the prisons and military, there is a place for private spaceflight orgs providing other private organizations with launch vehicles. But government....no way.
I'm unclear why it's okay for NASA to pay Boeing/Rockwell for an orbiter and Thiokol/Allianet for boosters (aka how they've been operating the manned space program since the 80s), but if they go with one company (SpaceX) for an entire system instead, this is suddenly a bad idea?
The problem with studies like this is, people look at the results and conclude, "So, if everyone used their real name in the Internet, it wouldn't be a problem." Try again. The same deindividualization would still occur. John Andersen is just as effectively anonymous whether he posts as John Andersen or TrollFace23. Uniquia McRarename may have a problem, but the remaining 98% of the forum users won't.
So are they really "jerks" or just being brutally honest and you can't handle it?
They are really jerks, and morons as well if they can't figure out how to be honest without being jerks.
What I find most interesting is that nobody has mentioned Diaspora since Google+ launched.
That's because nobody mentioned Diaspora before Google+ was launched, either.
In response to your comment, I have now mentioned it, but I honestly have no idea what we're talking about. I know what "diaspora" means, of course, but the only time I've ever seen it used with a big-D is when referring to the Jewish diaspora following the Babylonian exile. From the context in which you're using it, I'm guessing it's the name of some other social networking product that most people have never even heard of?
That means it's sure to win. This reminds of of when OS/2 mopped the floor with Windows because it had superior multitasking and memory management!
If only they'd had a superior product... or at least a decent UI. Lacking that, all the wonderful multitasking and memory management in the world was pretty useless.