glanced through my new socioeconomic Twitter analytics after @vogon and @0xabad1dea started tweeting about it and dang y'all are yuppies
@FioraAeterna @vogon @0xabad1dea yeah I've been wanting to look at how off their gender prediction algorithm is
-
-
@sarahjeong@FioraAeterna@0xabad1dea the big thing, I think, is that it's the percentage of people whose gender it can guess -
@sarahjeong@FioraAeterna@0xabad1dea and I'm sure sexism drives non-men to make their gender less marked in a bunch of ways -
@vogon@FioraAeterna@0xabad1dea right, but otoh I'm pretty sure their training data is based on people whose FB accounts are linked -
@vogon@FioraAeterna@0xabad1dea like what the fuck does that say about gender prediction skew -
@sarahjeong@vogon@0xabad1dea my guess is it's a really naive algorithm based on interests/etc which is actually really ironic -
@FioraAeterna@sarahjeong@0xabad1dea my instinct is that it's gendered display names + vocabulary analysis -
@vogon@FioraAeterna@0xabad1dea Yeah I mean it's got to be that also but they've got to be checking up against training data -
@sarahjeong@FioraAeterna@0xabad1dea I'd expect their training data is people with blue checkmarks - 2 réponses de plus
Nouvelle conversation -
-
-
@sarahjeong@vogon@0xabad1dea I tested a while back when they still let one-follower accounts have analytics; it thinks I'm male -
@FioraAeterna@sarahjeong@vogon@0xabad1dea Saaaame. I blame belonging to the computering lifestyle.
Fin de la conversation
Nouvelle conversation -
Le chargement semble prendre du temps.
Twitter est peut-être en surcapacité ou rencontre momentanément un incident. Réessayez ou rendez-vous sur la page Twitter Status pour plus d'informations.