Unhappiness on Instagram: Can we train algorithms to detect it?

unhappiness, social media

Researchers are developing algorithms to detect unhappiness on social media, which identify the basic needs of users from the content they share

Looking at mental illness, comparison, and the dangers of social media, researchers look for a solution to identify unhappiness on Instagram through a poster’s content – trained algorithms.

According to William Glasser’s Choice Theory, there are five basic needs that are central to all human behaviour: Survival, Power, Freedom, Belonging and Fun.

Researchers note that these needs even have an influence on the images we choose to upload to our Instagram page. Though this is a culturally normative thing for many people to do, social media can often hide mental illness or unhappiness for people online.

Mohammad Mahdi Dehshibi, who led this study within the AI for Human Well-being (AIWELL) group, said: “How we present ourselves on social media can provide useful information about behaviours, personalities, perspectives, motives and needs.”

In training algorithms, researchers aim to identify those who could be struggling with unhappiness based on the content of the images they post.

This work could potentially improve preventive measures for people suffering from unhappiness – which is commonly increased by social media usage – ranging from identification to improved treatment when a person has been diagnosed with a mental health disorder.

Using a deep learning model that identifies the five needs described by Glasser

The research team spent two years working on a deep learning model that identifies the five needs described by Glasser, using multimodal data such as images, text, biography and geolocation.

For the study, they analysed 86 Instagram profiles, in both Spanish and Persian.

Drawing on neural networks and databases, the experts trained an algorithm to identify the content of the images and to categorise textual content by assigning different labels proposed by psychologists, who compared the results with a database containing over 30,000 images, captions, and comments.

According to the researchers, the experiments “show promising accuracy and complementary information between visual and textual cues”.

They found that Spanish-speaking users are more likely to mention relationship problems

Glasser’s theory says that each choice we make does not respond to just one basic need, so the multi-label approach of this study is useful in clearing up this doubt.

Dehshibi, a research scientist at Universidad Carlos III de Madrid, uses an example to explain this: “Imagine that a cyclist is riding up a mountain, and at the top, they can choose between sharing a selfie and a group photo.

“If they choose the selfie, we perceive a need for Power, but if they choose the other option, we can conclude that the person is not only looking for fun but also a way to satisfy their need for belonging”.

The profiles analysed belonged to people who communicated in two different languages, to avoid cultural bias, as previous studies found, for example, that Spanish-speaking users are more likely to mention relationship problems when they are feeling depressed than English speakers.

The researchers said: “Studying data from social networks that belong to non-English speaking users could help build inclusive and diverse tools and models for addressing mental health problems in people with diverse cultural or linguistic backgrounds.”


Please enter your comment!
Please enter your name here