Translated data: New South Wales University researchers utilized data from a longitudinal study in Australia, focusing on adolescents aged 14-17, to collect over 4,000 potential risk factors for suicide and self-harm. By applying machine learning models to these factors, the researchers found that mental health conditions such as depression, anxiety, and behavioral issues were the most significant predictors of risk. This model, which considers a broader range of factors compared to just past suicide attempts, can more accurately predict future risks of suicide and self-harm. The study also identified school and family environments as significant influencing factors. This research demonstrates that using big data and machine learning can more precisely assess the risk of adolescent suicide, enabling early intervention. However, the application of the model must consider various social environmental factors and should not rely solely on individual psychological conditions.