STOP, artificial intelligence to prevent suicide
Research project to prevent suicide through artificial intelligence applied to social networks
According to the latest data from the INE, 4,227 people died by suicide in Spain in 2022. It is the highest number ever recorded in the country. One of the most concerning issues is the increase in suicide among adolescents, as cases have significantly risen from 53 to 75 compared to 2021. According to the WHO, each suicide has a severe emotional impact, at least on six people in the victim’s environment.
STOP is a research project that studies mental health problems on social networks through Artificial Intelligence. The goal is to find patterns related to the high risk of suicide and other disorders that can lead to this issue, such as depression or eating disorders. When a user matching a risk profile is detected, a targeted prevention campaign is launched, including information like the Helpline, available 24 hours a day, and other suicide prevention hotlines.
Some indicators taken into account include the times at which the person publishes content, interactions received from other users, their interests, and the type of posts. The team, consisting of psychologists, engineers, and psychiatrists, has developed algorithms that detect suicidal behavior patterns with 85% accuracy through texts, images, and social activities on Twitter.
Characteristics of innovation
Localization
Barcelona, Catalonia
Partners / Funders
UAB, Hospital Parc Taulí de Sabadell, and the Aid and Hope Foundation.
Genesis
STOP was born in early 2017. The driving force behind the project, Ana Freire, a researcher at the Department of Information and Communication Technologies at UPF, encountered a case on Facebook that inspired her to create the platform. A woman shared a farewell letter on the social network, where she had previously shown warning signs, and later took her own life. This case led Freire to investigate and carry out a project for prevention and intervention on social networks, as there were few initiatives in this regard.
Level of implementation
In 2020, a targeted campaign was conducted by STOP during the Christmas season through Facebook and Instagram. Over 24 days, it reached 667,655 people, with the highest number of views on January 1st: 90,993. According to the project organization, calls to the suicide prevention hotline from social networks have increased by 60%.
Banc d’innovacions