Soon, there's no point you create a profile and tweet are anonymous on the microblogging site Twitter. A group of researchers has found an algorithm that is able to guess the gender of 140 Twitter users simply tweet of his character.
This capability will help Twitter users from the pitfalls of other users that displays a fake profile. For example, an old man who pretends to be a lesbian women, as is the case in Damascus, when a middle-aged man from Georgia to make a fake account called "Gay Girl".
This study departs from the idea that women could wear a different language from men. For example, women usually prefer to use smileys and repeating letters.
The study was conducted by a group of researchers from MITRE Corporation, a technology research firm in Washington, the United States. Algorithm that accurately guess the sex of Twitter users with a way to isolate some specific words in each tweet.
Wow!Twitter can distinguish the sexes more accurately
Twitter does not display the sexes users. This is a perfect test for the scientific method with special algorithms.
The research team initially collects location, description, name the profile, and the real names of all Twitter users in their sample. Almost all samples of posting only once.
Opening tests done to see if the algorithm can detect the sex of a person's name and a computer capable of correctly guessed 89 percent.
By analyzing only the content of a single tweet, the algorithm was able to guess the gender correctly almost 66 percent. By analyzing all the tweets, its accuracy increases to more than 75 percent.
Another result is 71 percent accurate in the description only and 77 percent accurate at the name on the screen. When combining all four elements in the tweet, the level of computer accuracy to 92 percent.
Punctuation marks that appear often be an indication of gender. Use of smileys or interjection typically indicates a Twitter account owners are women.
Women seem to prefer to use words like love, cute, happy, mommy, sleep, school, baby, bed, chocolate, and hate. It also includes common abbreviations, such as "LOL" and "OMG".
While he has few phrases that became its attributes, including "http" and "google".
The study also showed that the sexes look at "bigram" possessive, that is the phrase that begins with my or our words. Phrases that become the attributes of men is my wife, my gf, and my beer. The women usually use my yogurt and my husband.
This algorithm is useful for anyone intending to reach specific audiences in the microblogging site, such as brand owners and businesses who want to market products to users of Twitter. But it must be useful also for those who do not want to get stuck
