It's said that the eyes are the windows to the soul, and they can sometimes reveal what people are thinking and how they feel.
Now artificial intelligence has come up with a detailed breakdown of someone's character by mapping tiny eye movements against a list of personality traits.
A study, by the University of South Australia, tracked the eye movements of 42 participants as they went about their daily lives doing ordinary tasks around a university campus.
The recorded movements were then compared to the result of personality questionnaires completed by the subjects.
Next, researchers used machine learning algorithms to map individual's eye movements with the 'big five' personality traits – openness to experience, conscientiousness, extroversion, agreeableness and neuroticism.
They found that each individual's unique pattern of eye movements could be linked to their personality traits.
By analysing the slight flickering of their eyes, machines could reliably assess what each personality was like – recognising four of the big five traits, with the exception of openness to experience.
Dr Tobias Loetscher, a researcher on the project, said the findings could improve interactions between humans and machines.
"People are always looking for improved, personalised services. However, today's robots and computers are not socially aware, so they cannot adapt to non-verbal cues," he said.
"This research provides opportunities to develop robots and computers so that they can become more natural, and better at interpreting human social signals."
More from Science & Tech
The research could be particularly useful because it was conducted in an everyday setting and not a laboratory, so it is likely to have recorded more natural behaviour than other tests.
"Thanks to our machine-learning approach, we not only validate the role of personality in explaining eye movement in everyday life, but also reveal new eye movement characteristics as predictors of personality traits," Dr Loetscher said.
[contf] [contfnew] 
Sky News
[contfnewc] [contfnewc]
