Crowdsourced Twitter Contest Aims To Predict Psychopaths From Their Tweets

Can you predict whether someone is a psychopath in 140 characters or less? The organizers of this crowdsourced data science contest think so!

Hosted on Australian crowdsourcing platform Kaggle, the “Psychopathy Prediction Based on Twitter Usage” contest asks data scientists to come up with an algorithm to determine someone’s personality – specifically whether or not they have a high degree of psychopathy – base on their Twitter use. The contest is organized by the The Online Privacy Foundation.

The organizers provided interested participants with a dataset of users’ self-assessed psychopathy scores along with 337 variables related to Twitter. They asked participants to play with the data and come up with a model that can predict the psychopathy score based on the Twitter data.

The contest is now closed, running from May 14 to June 29. A total of 1071 entries were submitted by 113 teams of data scientists and researchers, some with promising results.

No formal results have been posted yet.

The study is part of the “Big Five” factors of personality, which are openness to experience, conscientiousness, extraverstion, agreeableness and neuroticism. The 3,078 volunteers answered questions related to these five personality factors as well as others such as emotional stability and narcissism, in order to determine their degree of psychopathy.

Only 3 percent of the world’s population is considered psychopathic.

A similar study undertaken by The Online Privacy Foundation in 2011 found that making assumptions about someone’s personality based on their Facebook posts was wrong more often than it was right.

However, researchers at Cornell and the University of British Columbia found that certain words used frequently in social media – such as “since” and “um” – were often related to psychopathic behavior.

(Crazy computer guy image via Shutterstock)