A recent study published in Scientific Reports revealed that facial recognition technology can accurately predict someone’s political stance from their Facebook profile photo. According to the study, authored by Michal Kosinski, the algorithm shows greater accuracy in deducing a person’s political orientation than either human judgement or a personality test. A common facial recognition algorithm was applied to images of more than a million individuals drawn from their entries on the social media platform to predict their political orientation by comparing their similarity to faces of liberal and conservative others.
Kosinski, who is an associate professor at Stanford University, said that he has been studying privacy risks over algorithms over the last decade. While speaking to PsyPost, he said that companies collect data and develop algorithms aimed at extracting insights, but are reluctant to reveal how accurate such models are. Kosinski added that facial technology can accurately expose traits ranging from sexual orientation to personality.
Kosinski’s study applied a facial recognition algorithm to nearly 1,085,795 faces. Of this dataset, 977,777 came from dating website users in the US, UK and Canada who had self-reported their political orientation. The other 108,018 faces were from Facebook users in the US who had also self-reported their political orientation and additionally completed a 100-item personality test.
The algorithm then compared each participant’s facial features to the average facial features of liberals and conservatives. As per the study, the technology used these similarity measurements to determine the likelihood that a participant was either a conservative or a liberal. The result showed that the algorithm was able to predict political orientation alarmingly well and with similar accuracy across countries and social media platforms.
Accuracy over 70%
The study said that among US Facebook users, this accuracy hit 73 per cent and among US dating website users, accuracy was 72 per cent. Among dating website users in the UK and Canada, accuracy reached 70 per cent and 71 per cent respectively. The author of the study noted that the algorithm performed substantially better than humans, who are only able to distinguish between a liberal and conservative with 55 per cent accuracy.
Further, the study revealed that political orientation is linked to certain demographic traits that can be easily observed in the face. The researcher even explored whether certain features were tied to political orientation, including facial expression, eyewear, facial hair and head pose. Kosinski found that head orientation showed 58 per cent predictive power, with Liberals being more likely to face the camera directly. The study noted that emotional expression had 57 per cent predictive power, with Liberals being more likely to show surprise and less likely to show disgust. Kosinski said that his findings likely underestimate the intelligence of such technology, further adding that a higher accuracy would likely be revealed with higher resolution images, multiple images per person, or an algorithm specifically built for identifying political orientation.
Source: Republic World