<![CDATA[Mental health-related concerns have become increasingly visible in digital environments, where online search behavior reflects public awareness and information-seeking patterns. Examining temporal changes in mental health-related searches may provide valuable insights into emerging public concerns and support evidence-based planning. This study analyzes monthly Google Trends data on counseling and related mental health topics from January 2011 to June 2026. An XGBoost regression model was developed using historical counseling search interest, lagged variables, and related mental health indicators, including depression, anxiety, stress, suicide, and mental health search trends. Temporal analysis, feature importance evaluation, and SHAP interpretation were applied to investigate trend patterns and explain model predictions. The findings revealed a substantial increase in counseling-related search interest, particularly after 2018, with several periods showing notable fluctuations. The XGBoost model demonstrated satisfactory predictive performance, achieving a mean absolute percentage error of 18.78 percent and a coefficient of determination of 0.658. Historical counseling search variables were identified as the most influential predictors, indicating that previous search behavior plays a dominant role in forecasting future trends. The study highlights the potential of interpretable machine learning approaches for monitoring digital mental health trends and supporting data-driven public health strategies.]]>
Copyrights © 2026
