Facebook 'Likes' predict personality study

Occurred: May 2013

A research study revealed that Facebook Likes can accurately predict various personal attributes and personality traits, raising concerns about 

University of Cambridge and Stanford University researchers drew on data showing the Facebook likes of 58,000 volunteers in the US to design a set of algorithms to show how an accurate portrait of someone's personality, including their religious views, political beliefs, race, and sexual orientation could be made from the things they have 'liked' on Facebook.

The algorithms proved 88 percent accurate in determining male sexuality, 95 percent accurate in distinguishing African-American from Caucasian-American, and 85 percent in differentiating Republican from Democrat.

Digital rights and rivacy campaigners expressed their concerns about the study's findings, recommending people consider carefully what they choose to share online. 

Michael Kosinski, lead researcher on the project, concurred. "I can imagine situations in which the same data and technology is used to predict political views or sexual orientation, posing threats to freedom or even life," he said. 

System 🤖

Operator: Meta/Facebook
Developer: Kosinski M., Stillwell D., Graepel T.
Country: USA
Sector: Research/academia; Technology
Purpose: Predict personality
Technology: Machine learning
Issue: Privacy

Research, advocacy 🧮

News, commentary, analysis 🗞️