In Korea, where suicide is the main cause of death in the age of low birth rate and aging population, 90.1% prediction accuracy is achieved through interdisciplinary research using artificial intelligence and big data

Professor Kim Kyung-won, Department of Trade, Incheon National University   

A research paper was published by Professor Kim Kyung-won of Incheon National University and Professor Shin Soo-min of Yuhan University, which can predict high-risk groups who think of suicide in real time with a high accuracy of 90.1 percent by entering data from Korea's child and adolescent survey. The paper has been published in the "Asian Journal of Psychiatry," an international journal of SCIE with global authority in the top 5% in the field of Psychiatry.

Professor Shin Soo-min of Yuhan University

Korea has the highest suicide rate among the Organization for Economic Cooperation and Development (OECD) countries, and at the same time, it is experiencing the problem of low birth rate and aging population. In 2022, the suicide rate in Korea is 25.7 (the number of deaths per 100,000 people), and suicide is the number one cause of death for teenagers aged 10 to 19, and the suicide rate for teenage men is 18.8%, which is the steepest compared to other age groups. The Republic of Korea is implementing various suicide prevention projects in cooperation with the central government, local governments, and private organizations to actively cope with and respond to the suicide problem. Since 2004, a basic plan for suicide prevention has been established and implemented, and the government has included "suicide prevention and the spread of a culture of respect for life" as one of the national tasks. It is also increasing the linkage rate of various social services in the community for suicide attempters and strengthening the follow-up management of emergency rooms. In 2022, the Suicide Prevention Act was revised to prevent suicide and create a culture of respect for life.

According to 2020 statistics, teenage suicide motives were reported to be mental difficulties for both men and women. Suicide thought is one of the important precursors to suicidal planning or attempts, and 15.6% of those who think of suicide try to commit suicide within a year, and 31.8% try to commit suicide at least once in their lives. Therefore, if you can quickly predict and respond to high-risk suicide groups, you can contribute to actively preventing suicide.

Machine learning or deep learning, which is the basis of artificial intelligence, has been used to predict diagnosis and disease in health and medical fields, but in social science, big data-based artificial intelligence analysis is still somewhat unfamiliar and poorly utilized. Therefore, by presenting a big data-based analysis method that can quickly and effectively predict suicidal thoughts, this study expanded the horizon of the methodology for deriving a data-driven suicide prevention policy and contributed to the scientific identification of the causes of suicidal thoughts.

After learning big data from 2017 to 2020 collected by the Korea Youth Policy Institute with machine learning and deep learning algorithms, which are the basis of artificial intelligence, a model reflecting all relevance patterns related to suicidal thoughts was created and applied to future data, it was possible to predict the suicidal thoughts of children and adolescents in real time with more than 90% accuracy. In addition, it was confirmed that children and adolescents with sadness and depression had more than 25 times suicidal thoughts than those who did not, and children and adolescents with anxiety, loneliness, and abusive experiences had more than seven times suicidal thoughts.

This approach has shown very promising results in AI identifying signs of suicidal thoughts in children and adolescents and predicting high-risk children and adolescents in the future, and has led to the spread of a data-driven culture that solves social problems through convergence or interdisciplinary research from a social science perspective.

Professor Kim Kyung-won, who conducted the study, said, "Artificial intelligence is being quickly applied to various businesses and helps make many decisions in our lives, but it has not been used much in areas for the state to operate as an effective social safety net. This study can contribute to rapidly predicting and responding to suicide of other age groups, such as ordinary adults and the elderly, as well as children and adolescents, who are the key players of future generations, and has the potential to be used in areas requiring social security by quantitatively presenting the causes of suicide that change in real time."

Rapper Logic's song '1-800-273-8255', released in 2017, is a title derived from the actual counseling hotline number of the National Suicide Prevention Lifeline in the United States. After two performances at the MTV Video Music Awards, the most media-exposed network, on Aug. 27, 2017, and the Grammy Awards, the most prestigious music awards ceremony, on Jan. 28, 2018, the number of suicide-related counseling calls to anti-suicide centers surged by thousands and the number of suicides that occurred decreased significantly. It is also important to design effective government initiatives and implement public-private partnerships to prevent suicide, but it is also time to consider how to use suicide prevention campaigns or cultural and artistic content that can actually reach children and adolescents. Identifying children and adolescents' suicidal thoughts early and providing practical responses and support in a timely manner also have important implications for risk prediction.