The content published on social networks can be analyzed to draw interesting conclusions about the different communities.

Both the personality of the people, as well as their needs and perspectives, are exposed for those who know how to analyze the material, as well as the feelings when sharing an image, a video or any other publication.

Researchers at the UOC in Catalonia have been working on this, developing an algorithm that can identify whether or not people are happy just by analyzing what they share on social networks. The goal is to be able to diagnose possible communication and mental health problems.

The deep learning model has been in development for two years, using the theory of American psychiatrist William Glasser, which describes five basic needs fundamental to all human behavior: survival, power, freedom, belonging and fun. Those needs are what affect when choosing what to upload to social media platforms, such as Facebook, Twitter or Instagram.

Among the first conclusions we have something interesting among Spanish-speaking users: they are more likely to mention relationship problems on social networks when they feel depressed.

86 Instagram profiles, both in Spanish and Persian, were analyzed for the study published in the IEEE Transactions on Affective Computing journal. They believe their research can help improve preventive measures by identifying and treating those who may have a mental health disorder.

As an example, they indicate that if a person climbs a mountain and takes a selfie, it is perceived as a need for power, while a group photo indicates a way to satisfy their need to belong to a group.