Node: A Real Time Smartphone Big Data Application for Health Care Epidemiology

Abstract

In this thesis the author outlines the development of a smartphone application from design to deployment in order to facilitate a more comprehensive and real time understanding of the spread of a disease in a population. Particular to this study an application was developed which does not rely on users using a specific hardware model of Android headset, but rather is only restricted to software (Jelly Bean or higher). This presents a novel step forward over previous research done, which is classically limited to a few smartphone users or key chained devices. Furthermore, this thesis demonstrates the coupling of three vectors of data, in this case survey, location specific and documented health cases. These three vectors are then combined into one health profile referred to as the Node Network, which presents a user's "health status" as a combination of these traits and includes active risk analysis based on their projected path and movements throughout the day. By analyzing a user's accelerometer data this study additionally allows for a medical practitioner, policy maker, or academician to study and correlate a user's behavior to their health state. Additional details provided via GPS location such as where a user eats, the type of food a user is eating, etc. which have never been studied in previous studies, are presented in this study. The successful coupling of these vectors allows for an unprecedented insight into civilian lifestyles which has huge impacts for health care providers and government officials in both developing and developed countries. It is the intention of the author that this study serve as the beginning of the development of a health forecasting system, which will allow interested parties to forecast and predict and prevent biological attacks, epidemic outbreaks, and result in better health policy initiatives specific society.