How artificial intelligence could help detect burnout

BURNOUT is an increasingly important subject at a time when a part of the population seems to be close to professional exhaustion. A new method based on artificial intelligence and text analysis could help diagnose the condition more efficiently.

Could artificial intelligence help facilitate the diagnosis of burnout?

While many tools and tests exist online to try to address the issue, a new method, based on artificial intelligence and the use of natural language processing, could offer new perspectives.

NLP (Natural Language Processing) is a technology that involves automatically analyzing sentences formulated by humans in order to make a decision or identify a behavior.

In order to detect signs of burnout in texts using this kind of model, a large amount of data needed to be accumulated and analyzed. This was achieved through the Reddit platform.

The data gathering phase involved storing anonymous texts about all kinds of experiences with no less than 13,568 samples. In this mass of stories, 352 were related to burnout and 979 to depression.

The objective for the model is to succeed in automatically detecting which statements are related to burnout. This method achieved a success rate of about 93% in identifying burnout cases.

"Natural language processing effectively detects burnout and does it relatively efficiently, which is very promising," explains Mascha Kurpicz-Briki, professor of data engineering at the Bern University of Applied Sciences in Biel, Switzerland, who oversaw the project.

A tool to support health professionals

According to the French National Authority for Health (HAS), burnout is defined as physical, emotional and mental exhaustion resulting from prolonged investment in emotionally demanding work situations.

The main symptoms can be physical, emotional or even cognitive, with the onset of sleep disorders, anxiety or emotional fatigue.

The World Health Organization characterizes the condition by feelings of low energy or exhaustion; increased mental distance from a job, or feelings of negativism or cynicism relating to that job; and reduced professional efficacy.

As promising as the results from this text-analyzing AI model may be, human input should not be overlooked, according to the study authors.

Instead of replacing health professionals and mental health specialists, this technology should no doubt remain an additional support tool to aid their decision-making.

To further verify the model's results, the next step will be to use this AI system on real cases in a representative sample of the population rather than on anonymous testimonies.