The things you post on social media can sometimes say a lot about you as a person, and even your mental state, and this is something that researchers are looking to explore more of. In a recently published study in the EPJ Data Science (via Mashable), researchers used machine learning to try and detect signs of depression from Instagram accounts.
The system used a variety of factors to make its decision, such as choice of color, the filters used, face detection, user comments, and how much one is engaged in their posts. The research suggested that users who post bluer, darker, and grayer images tended to be more depressed, and were also less likely to use Instagram’s filter, or if they did, would choose the black-and-white Inkwell filter.
In contrast, those who did not show signs of depression tend to use filters such as Valencia, which seems to be in line with a research from last year that also explored how photos posted onto social media could be used to detect depression in a person. However it has been pointed out that there is a caveat with this study, and that is it has a very small sample size and lack of demographic information.
This means that while there appears to be some correlation, there isn’t quite enough to go on just yet, but it could be the start of something. According to Chris Danforth, co-author of the study and Flint Professor of Mathematical, Natural, and Technical Sciences at the University of Vermont, “Doctors don’t have visibility into our lives the way our mobile phone does. It knows a lot more about us than we know about ourselves.”