Facebook knows you better than your friends do - because Likes reveal so much about your character

Study of 86,000 users reveals the power of intelligent machines, and they're getting better

The study involved using machine-learning software to analyse Facebook “likes” to assess a person’s personality based on the five main traits commonly used in psychological assessments: openness, conscientiousness, extraversion, agreeableness, and neuroti (Getty)

A computer can be better at assessing someone’s basic personality than close friends or family and is almost as good as someone’s spouse, scientists have found.

A study of more than 86,000 users of Facebook has demonstrated the power of intelligent machines to predict an individual’s character based on what they have listed as their Likes.

Researchers said that the day when computers are able to judge a person’s personality accurately has almost arrived and even suggested that science fiction films like Her, based on a man’s emotional attachment to an intelligent computer, are closer than we think.

“In the future, computers could be able to infer our psychological traits and react accordingly, leading to the emergence of emotionally-intelligent and socially-skilled machines,” said Wu Youyou of Cambridge University.

“In this context, the human-computer interaction depicted in science fiction films such as Her seem to be within our reach,” said Ms Wu, the lead author of the study published in the journal Proceedings of the National Academy of Sciences.

“People may choose to augment their own intuitions and judgements with this kind of data analysis when making important life decisions such as choosing activities, career paths or even romantic partners,” she said.

In Her, starring Joaquin Phoenix, a man develops a close relationship with an artificial intelligence system called Samantha, communicated through a female voice, which seems to understand him.

In the Cambridge study, the scientists used machine-learning software that analysed Facebook “likes” to assess a person’s personality based on the five main traits commonly used in psychological assessments: openness, conscientiousness, extraversion, agreeableness, and neuroticism.

For instance, liking the film Kite Runner, mountain biking or studying indicated that a person is well organised and conscientious; while liking the Velvet Undergrond, Blackadder or Douglas Adams suggest they were more spontaneous and less conscientious.

The more “likes” the computer could include in its assessment, the more accurate it became in forming a prediction of a person’s overall character and personality, said David Stillwell of Cambridge, who developed the personality test.

“We used some machine learning technology to see how accurately we could predict personality and we compared this set of predictions made by the computer with the set of predictions made by friends and family, and by the person themselves in a self-assessment,” Dr Stillwell said.

'Her' tells the story of a man who develops a close relationship with an artificial intelligence system called Samantha (Rex)

“We found that the computer can predict personality as accurately as the person’s spouse, better than a close friend or family member and a whole lot better than a work colleague,” he said.

For instance, the computer was able to predict an individual’s personality more accurately than a work colleague by analysing just ten “likes”, such as films, books, sports or celebrities.

The computer did better than a friend or roommate after analysing 70 likes and was better than a parent or sibling after assessing 150 likes. The machine could even predict personality almost as well as a husband, wife or partner after analysing 300 Facebook likes, the researchers found.

Given that a typical Facebook user lists about 227 “likes”, and this number is steadily rising, the researchers believe this kind of computer assessment has the potential to know us better than close friends and family.

They also suggest that the computer assessment of peoples’ personalities has a wide range of potential uses, which could lead to ethical questions about how and when it is applied.

“The ability to judge personality is an essential component of social living, from day-to-day decisions to long-term plans such as whom to marry, trust, hire or elect as president,” Dr Stillwell said.

“The results of such data analysis can be very useful in aiding people when making decisions…People need to be aware of it, and companies need to tell us when they are using it,” he said.

Dire traits? The Facebook personality

People with certain personality traits are more likely to “like” certain films, artists, activities and phrases. These traits include:

OPENNESS

Liberal & Artistic (Buddhism, David Bowie, Bauhaus)

Conservative & Conventional (George W Bush, Fox News, Deal or No Deal)

CONSCIENTIOUSNESS

Well organised (Mountain biking, The Apprentice, study)

Spontaneous (The Velvet Underground, The Mighty Boosh, Blackadder)

EXTRAVERSION

Outgoing & active (Gucci shoes/clothing, tanning, meeting new people)

Shy & reserved (Doctor Who, Wikipedia, the game Minecraft)

AGREEABLENESS

Cooperative (Christian, Life of Pi, The Bourne Identity)

Competitive (Richard Dawkins, WikiLeaks, Lolita, Silence of the Lambs)

NEUROTICISM

Emotionally unstable (Kurt Cobain, I miss you, Gothic rock, Buffy the Vampire Slayer)

Calm & relaxed (Snowboarding, volunteering, ping-pong, kayaking)

Dire traits? the facebook personality