SL-Japan research collaboration leverages explainable AI to strengthen public health through a landmark international research project that demonstrates how artificial intelligence can support policymakers in understanding the complex social and economic factors behind suicide and homicide.
SL-Japan research collaboration leverages explainable AI to strengthen public health through an international study on suicide and homicide trends in the Americas
The study, titled “Explainable AI for Public Health Surveillance: Investigating the Persistent Crisis of Intentional Injury Mortality (Suicide and Homicide) in the Americas,” was recently published in the prestigious journal Scientific Reports. It brings together researchers from Sri Lanka and Japan, highlighting the growing contribution of Sri Lankan academics to globally significant research in artificial intelligence, public health, and data science.
The international research team comprises Sherin Kularathne, Dr. Namal Rathnayake, Prof. Ruwan Jayathilaka, Prof. Iori Nakaoka, and Prof. Yukinobu Hoshino. Their collaboration reflects the increasing importance of cross-border academic partnerships in addressing global health challenges through advanced analytical technologies.
The publication marks a significant milestone in Sri Lanka Japan research cooperation. Sherin Kularathne, a former MBA student at the Sri Lanka Institute of Information Technology (SLIIT) and currently a PhD candidate at Kochi University of Technology in Japan, worked alongside Prof. Yukinobu Hoshino from the university’s School of Data and Innovation.
The Sri Lankan contribution extends further through Dr. Namal Rathnayake, who is attached to the Advanced Institute for Marine Ecosystem Change in Yokohama, Japan, while Prof. Ruwan Jayathilaka represents the SLIIT Business School. Prof. Iori Nakaoka, meanwhile, is affiliated with the Faculty of Data Science at Shimonoseki City University in Japan.
The researchers analysed data covering 25 countries across the Americas over a 20-year period from 2000 to 2019. Their investigation focused on a range of socioeconomic and governance-related indicators, including unemployment, inflation, economic growth, and perceptions of corruption, to better understand the underlying drivers of intentional injury mortality.
Unlike conventional artificial intelligence models that primarily generate predictions, the researchers employed explainable AI techniques to make the decision-making process behind those predictions transparent. This enables researchers, healthcare professionals, and policymakers to identify the key variables influencing public health outcomes rather than relying solely on statistical forecasts.
The study combined snapshot modelling with persistence-aware modelling, enabling researchers to examine both current conditions and long-term historical patterns. The findings revealed that persistent structural socioeconomic conditions have a substantial influence on suicide and homicide rates, suggesting that sustained policy interventions are essential for achieving lasting improvements in public health.
Commenting on the findings, Prof. Ruwan Jayathilaka emphasised the broader role of artificial intelligence in evidence-based policymaking.
“Artificial intelligence should not be used only to predict outcomes. It must also help researchers and policymakers understand why those outcomes occur. Explainable AI allows us to identify the factors that influence public health crises and support more informed decision-making.”
A key feature of the research was the application of SHAP (SHapley Additive exPlanations) analysis, an advanced explainable AI technique that interprets the contribution of individual variables within machine learning models. By using SHAP analysis, the researchers were able to provide policymakers with clearer insights into how different economic and governance factors influence intentional injury mortality, making public health surveillance more transparent, targeted, and evidence-based.
The study demonstrates the expanding role of explainable AI in addressing complex societal challenges beyond traditional healthcare applications. As governments increasingly rely on artificial intelligence to support policy development, explainable AI offers a practical framework for ensuring transparency, accountability, and greater confidence in AI-assisted decision-making.
Beyond its scientific contributions, the publication also showcases the growing international reputation of Sri Lankan researchers and higher education institutions in emerging technologies. By combining expertise in artificial intelligence, public health, economics, and data science, the Sri Lankan and Japanese research team has produced findings that could help shape future public health strategies not only in the Americas but also in other regions facing similar social challenges.
The research underscores how international collaboration, innovative data science methodologies, and socially responsible artificial intelligence can work together to generate practical insights for policymakers. As governments continue to seek evidence-based solutions to complex public health issues, studies such as this highlight the transformative potential of explainable AI in supporting more effective, transparent, and informed decision-making worldwide.