A new study found that Facebook could help predict users with depression.
Chances are you are one of the 2.2 billion active users on Facebook who scroll through endless amounts of memes, dog videos and status updates.
A new study published in the Proceedings of the National Academy of Sciences, analysed the language each consenting patient used to predict future diagnosis of depression. Using language that references typical symptoms like loneliness, sadness hostility, rumination, and increased self-reference can be predictive of depression. The study looked at 1200 consenting patients, 114 of whom had already been diagnosed with depression. When looking at status updates, comments, likes etc. prior to their diagnosis the machine learning algorithm was able to match the patients who had been diagnosed. The AI program was most effective in predicting depression six months prior to a patients actual diagnosis and with significant accuracy up to three months prior their diagnosis.
Mentions of loneliness or isolation, such as “alone”, “ugh” or “tears” are early warning signs; as well as the number and length of posts.
“There is a perception that using social media is not good for one’s mental health, but it may turn out to be an important tool for diagnosing, monitoring, and eventually treating it,” said H Andrew Schwartz, principal investigator of the study.
The algorithm was fed 524,292 Facebook status updates by scientists from both people suffering from depression and who are not. The algorithm was set up to detect a range of depression associated language markers – those which pointed to emotional and cognitive cues towards loneliness, sadness etc. It was then tested against status updates from those who are diagnosed with depression for commonalities.
“What people write in social media and online captures an aspect of life that’s very hard in medicine and research to access otherwise,” – Schwartz.
While it is only a small study and can be refined in many ways, it is a first step towards proving that phone data and actual behaviour on social media can perhaps paint a more comprehensive picture about a person’s life than simply what they say in a clinic.