A new Israeli development may save people from suicide. This is how it works - voila! health

Iris Cole interviews Dr. Oren Tana about depression and its treatment (Walla system)

Every year there are about 6,000 suicide attempts in Israel, about 500 of which end in death.

If these numbers sound unbelievable to you, you should know that there is an underreporting of about 23 percent, so they are probably even higher.

In light of this, researchers and professionals are trying to find ways to help people with suicidal thoughts in real time, and now it seems that they have taken a real step in this direction.

A new algorithm developed at Ben Gurion University analyzes mental support conversations via chat, combining psychological theories in the field of suicide and mental distress with advanced artificial intelligence algorithms for language and text processing.

The support calls are carried out by associations that provide online emotional support through chats, WhatsApp and online forums.

This is the first study conducted in the Hebrew language for the computerized identification of suicidal risk during online support conversations.

The algorithm analyzes a dialogue between an applicant and a professionally trained volunteer, and alerts in real time if the dialogue indicates a high level of suicide risk.

Texts that indicate severe mental distress will alert the volunteer to an increased suicidal risk, and accordingly the volunteer will decide whether to involve external factors such as an emergency drive and/or activate other external factors that will immediately save lives.

17,000 calls, 91 percent accuracy

"Among 5,000 men and 12,000 women in the age range of 10-65, the study showed over 91 percent of accuracy in identifying suicidal discourse according to the accepted index in the field," noted Prof. Kobi Gal, the leading researcher from the Department of Software and Information Systems Engineering at Ben Gurion University of the Negev.

The study was conducted in Israel through data collection and analysis of over 17,000 anonymous chat conversations collected over 5 years.

The conversations took place on the digital system of the Sahar (Online Assistance and Listening) association, which provides answers and online psychological assistance and works to prevent suicide. The research group combined researchers from the field of information systems and software engineering, language analysis and psychology, a collaboration that allows for the construction of an accurate algorithm that knows how to analyze the text and identify Suicidal thinking during the online conversation with the help of psychological tools.

"Adding a theoretical background from the field of psychology and working with the Sahar association allowed us to build much more accurate algorithms," explained researcher Dr. Avi Segal, from the Department of Software and Information Engineering at Ben Gurion University of the Negev. "This emphasizes the importance of interdisciplinary collaborations with researchers from other worlds of knowledge and with organizations operating in the field."

More in Walla!

6,000 Israelis try to commit suicide every year - and there is no one to help them

To the full article

Analyzing conversations in the Hebrew language is even more difficult.

Illustration of a depressed woman (Photo: ShutterStock)

"It is important to note that analyzing conversations in the Hebrew language is significantly more challenging than processing texts in English, and this is due to the morphological complexity of Hebrew," noted Dr. Oren Zur, head of the Center for Research in Digital Politics and Strategy and a researcher from the Department of Software and Information Systems Engineering at Ben Gurion University of the Negev. Other researchers who took

part In the study: Prof. Yossi Levy-Balez, from the Faculty of Social Sciences at the Rupin Academic Center, and master's students Amir Bialer and Daniel Izmailov from Ben-Gurion University of the Negev. The development of the model was published at the COLING 2022 conference.

• health

• news

Tags

• suicide

• depression

• Ben-Gurion University