A single message on Twitter can reveal more of what we want
Twitter is a public forum widely known by all, if tweeted something about us we would like to keep it private, but almost always we fail in the attempt. However, people do not realize that only use Twitter can reveal more than what we want.
Data specialists are becoming adept at when making predictions about people based on very subtle clues in social networks, such as the language used, the network of friends or topics about which users speak. Through finding patterns in endless streams of Twitter data, experts have developed ways to predict the age, sex, location, personality, political orientation and if the user is depressed, even if these data are not explicitly disclosed by users.
“You NOT JUST SAY WHAT YOU SAY OR HOW: YOUR FRIENDS AND FOLLOWERS CONNECTIONS ALSO WITH MUCH ABOUT IT REVEALED, SO EVEN IF YOU 100% SILENT, THERE IS STILL MUCH TO BE ABOUT IT INFERRING”
As he stated by Christo Wilson, a computer scientist at Northeastern University who studied the effects of the algorithms used to try those online consumers.
The new step in this direction, published in the September issue of the journal PLoS One, shows how it is possible to guess almost unequivocally the monthly income of a tuitero.
Related articles: 5 Reasons Social Media Make Your Best Ally
In the study, researchers followed nearly 5,000 real Twitter profiles that clearly described the work of the person, whether it be as an executive director or miner. Based on the classifications of government jobs in the UK, compared the work of the person seeking their average salary and Twitter usage patterns to help them predict every salary.
The data collected could be more than useful for marketers, employers and even government agencies surveillances
After creating the predictive model, researchers could estimate the salary of unknown users with great precision, suffice to say that a user was at 5% of income while another was at 20%.
The team conducted this study because they are interested in making more demographic data were available for social science research. However, we all know that such predictions could be more than useful to marketers, data brokers, employers and government watchdog agency.