Dengue propagation prediction using human mobility

Abstract

This paper discusses about predicting Dengue outbreaks in Sri Lanka using heterogeneous data sets: Mobile Network Big data and epidemiological data. Up to now, Dengue epidemiological prediction was largely done using the past Dengue cases and weather data. However, very recently it was discovered that infection can propagate through humans, where an infected human travels to a vulnerable area and mosquitoes of that area will bite him, and contract and spread the virus. Hence this research explores the potential of incorporating human mobility, derived through mobile network data in predicting Dengue propagation. This paper presents the various data sources, how the data fusion was conducted and how the fused data was fed into the model and the results obtained and a discussion thereof, including the potential of extending the research.