What does Facebook know: New app uses what you 'like' to predict your personality

Computer programme assess 'big five' personality traits and predicts details including user's intelligence, life satisfaction and sexual preference

Click to follow
The Independent Online

Researchers at Cambridge University have developed a computer programme that can produce a detailed psychological profile based on a user's Facebook 'likes'.

The application, called 'Apply Magic Sauce', estimates details including a user's intelligence, political and religious views, life satisfaction and sexual preference.

It also assess openness, conscientiousness, extraversion, agreeableness and neuroticism, seen by many psychologists as the 'big five' personality traits.

The programme would show users what Facebook really knows about them and might offer an indication as to what assumptions those delivering personalised advertising have made.

But while its assessment is said to be more accurate than predictions made by friends or colleagues, it appears it sometimes misses the mark - journalist Sophie Kleeman, writing about the API for .Mic, claimed it had assessed her as a contemplative 29-year-old man.

The application works by comparing a user's Facebook 'likes' to those of six million other people. To do so a user needs to grant the API access to their Facebook account.

Speaking to the Mail Online, Dr David Stillwell, the deputy director of Cambridge University's Psychometrics Centre and a researcher behind the project, said that part of the tools purpose was to show what data was being collected from online behaviour.

"One of the challenges with our modern online lives is that data is being collected about us and predictions are made, but we never get to see what Facebook is doing behind the scenes," the Mail Online quoted him as saying.

"We wanted to show people, given this fairly innocuous data, this is the prediction Facebook can make about them."

Apply Magic Sauce is also marketed to businesses as allowing for greater personalisation and offering a psychological assessment based on online behaviour.

Comments