A recent study claims to have found a way to tell when social media users are depressed. It seems these people tend to post more photos, many of them containing faces but few having filters applied. Scientists managed to find out about this after designing an algorithm to analyze photos and identify if the person who posted it is depressed.
This project was developed by researchers from Vermont University and Harvard University. They developed a complex program which can tell if a person is likely to be depressed just by analyzing the photos and posts they share on social media. This is incredibly useful, as the algorithm can trigger an alert for people who might be really having a mental health issue.
Those photos posted by depressed people shared certain elements
The study was published in the journal EPJ Data Science. Results showed how the computer program managed to accurately identify depression in 70 percent of cases. However, something interested was also noticed. Doctors had an accuracy of only 40 percent in diagnosing depression.
Researchers identified several common elements in all photos posted by depressed people. These photos were more likely to feature a face, but without any filters. Also, they were darker, and attracted more comments than other photos.
Since people are now spending so much time online and sharing so much of their lives, finding out how to identify depression just by looking at social media posts is a huge achievement. This might lead to the development of a special app which warns your doctor if it detects any changes in your behavior.
Initially, the program analyzed the photos posted by people who were already diagnosed with depression. Then, it used what it had learnt to predict who else might be suffering from the same condition. The sample researchers used might have been a bit small, but the differences in posts were clearly visible, so the study collected important information.
Image Source: Pixabay