Detecting mental disorders through social networks

What if social networks could be used to monitor the mental health of Internet users? This is the aim of the Canadian research project "Mood", in which two Montpellier researchers from the Laboratoire d'informatique, de robotique et de microélectronique de Montpellier (Université de Montpellier - CNRS) are participating.

Credit: UM - David Richard

Every minute on Facebook, 293,000 statuses are updated, 510,000 comments posted and 136,000 photos published. A veritable goldmine of data for researchers, and soon for doctors who will be able to track their patients suffering from mental disorders thanks to their social networking posts.
How will this work? "Thanks to text mining and discourse analysis tools that enable us to understand what Internet users are feeling by deciphering their publications," explain Sandra Bringay and Jérôme Azé, researchers at the Montpellier Laboratory of Computer Science, Robotics and Microelectronics (Lirmm, Université de Montpellier - CNRS). "We are developing algorithms that can, for example, identify emotions such as sadness or self-loathing in speech, to spot Internet users who are in a bad way", says Sandra Bringay.

Prevention 2.0

And what about privacy? "The Moodproject is not designed to identify specific Internet users; it processes anonymous data," reassures the researcher. On the other hand, it does enable us to better understand how people express themselves, and to test the algorithms developed by the researchers. " These results can be used by doctors, for example, to monitor the mental health of patients who have given their consent," explains Sandra Bringay.
"Using social networks to identify patients suffering from depression, anorexia or harassment will enable them to be better cared for", the researchers predict. These results could prove particularly useful in preventing suicide, which claims three times as many lives as road accidents every year.