Facebook decides you're 'gay' if you 'Like' something on Lady Gaga, says study

The assumptions would affect the image of LGBT people, who are not yet accepted by people openly and might provoke real-life risks against the community.

Facebook
Facebook Reuters

A new study has revealed that Facebook could judge a person's sexual orientation and the drug history of that individual based on what the person is sharing on the social media wall and their 'likes' on a particular content.

The study entitled "Enhancing Transparency and Control When Drawing Data-Driven Inferences About Individuals," is a joint research material by Daizhuo Chen from Columbia Business School, Northeastern University's Samuel Fraiberger, Robert Moakler and Foster Provost from Stern School of Business, New York University.

Although it is not a perfect way to judge someone's interest, the study claims that if a person hits 'like' on something related to Lady Gaga or Human Rights campaign, Facebook will characterize that person as gay.

This study has examined that how organizations use information about an individual to predict their characteristics and how those individuals could protect their privacy. The research has also introduced a "cloaking device."

According to the study "The cloaking device essentially tells the system: 'do not draw inferences like this about me'—or more practically, 'do not show me ads or content for the same reasons that you decided to show me this,'"

In the study, it is also mentioned that while using cloaking device, the analytical tools there will be two questions, (1) how much information must users cloak to significantly affect inferences about their personal traits? (2) Can firms change their modelling behaviour to make cloaking more difficult?

On the other hand, reports said that although there is no specific information about how Facebook is going to decide sexual orientation, the assumptions would affect the image of LGBT people, who are not yet accepted by people openly and might provoke real-life risks against the community.

"While some online users may benefit from being targeted based on inferences of their personal characteristics, others may find such inferences unsettling," said Daizhuo, Samuel and Robert.

Related topics : Facebook
READ MORE