A Framework for IoT Data Collection and Fusion in Infectious Diseases Surveillance

Abstract

Although IoT has proven to be a vital instrument for data collection, epidemiological data collection has presented a significant challenge to the AI community. Thus, data for epidemiological research from multiple sources needs to be integrated to gain a holistic insight into epidemiological incidences. Hence, the authors aim to provide a framework for collecting, aggregating, and fusing data from diverse IoT sources for epidemiological research. Thus, a wireless network is designed for energy efficiency and communication efficiency. In addition, the CoAP mechanism for data transfer to a cloud-based service for aligning, de-duplicating, aggregating, and mapping data from several sources is presented. Subsequently, the researchers trained three machine learning models to predict disease incidence, vector abundance, and host population using both synthesized data using the proposed framework and data captured using traditional means. They assessed the performance of the two models by measuring their accuracy and learning rate. Results show the superiority of the proposed framework.