Virtual Booth

Dengue fever remains a major public health concern in South Colabato, Philippines, where delayed detection and underreporting limit effective outbreak response. This study developed ASCLEPIUS, an Al-enhanced computational framework for real-time dengue outbreak geospatial analysis using mathematical modelling and machine leaming forecasting. Historical dengue data (2014 - 2024) from Tupi's 15 barangays, climate records (temperature and humidily) from MDRRMO and DOST, and barangay-level population data were analyzed, Multiple Linear Regression (MLR) and Long Short-Term Memory (LST) models were constructed and integrated into a web-and-mobile platform with geospatial mapping. symptom logging, and an Al chatbol for denguerelated recommendations. The LSTM model achieved superior predictive accuracy (R²=0.9543) compared to MLR (R²=0.9350). Projections for 2024 - 2050 indicated increasing dengue risks under climate variability. The chatbot achieved 98% accuracy, while notification speed analysis confirmed significant differences (p<0.05), with most alerts delivered within two seconds. User evaluations rated funclionality, usability, acceptability, and adaptability as highly functional. ASCLEPIUS demonstrates strong potential as a scalable tool for proactive dengue surveillance and public health preparedness.

Keywords: Dengue forecasting, Al chalbol, geospatial analysis, machine fearing, outbreak monitoring, mathematical modelling, regression model