Mining Enormous Mobile Datasets to Improve Mitigation Strategies for Limiting the Spread of Infectious Disease

We secured access to a dataset containing the entire anonymized call detail records (phone/text) of 5 million mobile phone subscribers in Cote d’Ivoire, tracked over a 5-month period. The goal of this work-in-progress is to analyze the dataset for information that could help public health officials develop more effective strategies for limiting the spread of infectious disease. Using antenna (cell tower) proximity data to situate subscribers, clustering algorithms were applied to identify groups of individuals expressing similar mobility patterns. Incorporating this knowledge of dynamic population densities could lead to better-informed quarantine/isolation decisions