Instagram is arguably the most popular photo sharing app online and has been an avenue for people to share their glamorous lives with the world but according to a recent study published in the EPJ Data Science journal, the shades of a shared image can be used to judge the person’s mental health.
Also on Guiding Tech
The researchers used a group of 166 people, out of whom 71 were diagnosed with depression and the rest were ‘healthy control group’ who were used to compare and also to draw a bench mark of normal behaviour.
A total of 43,950 Instagram image posts were analysed using colour analysis, metadata components and algorithmic face detection.
“In our research, we incorporated an ensemble of computational methods from machine learning, image processing, and other data-scientific disciplines to extract useful psychological indicators from photographic data. Our goal was to successfully identify and predict markers of depression in Instagram users’ posted photographs,” the researchers stated.
The study supported the following two of the three hypothesis put forward by the researchers:
Instagram posts made by individuals diagnosed with depression can be reliably distinguished from posts made by healthy controls, using only measures extracted computationally from posted photos and associated metadata.
Instagram posts made by depressed individuals prior to the date of first clinical diagnosis can be reliably distinguished from posts made by healthy controls.
“Our findings establish that visual social media data are amenable to analysis of effect using scalable, computational methods. One avenue for future research might integrate textual analysis of Instagram posts’ comments, captions, and tags,” the researchers concluded.
According to the research paper, the methodology was able to identify depression in people with 70 percent accuracy, which is greater than the average 50 percent accuracy of doctors while diagnosing depression.
The research is only a proof of concept right now and it can’t be said for sure whether it can have a similar application on a larger scale. Check out the study in detail at the EPJ Data Science.