Metadata capital: Simulating the predictive value of Self-Generated Health Information (SGHI)

Metadata is crucial for understanding data, and can be viewed as a form of capital in the context of Big data. This paper reports on research simulating the potential of SGHI (Self-Generated Health Information) for predicting asthma episodes. A data set of 2,000 cases was generated using the Monte Carlo simulation method, with secondary modifications on air quality and geo-location. The research is being pursued as part of a National Consortium for Data Science (NCDS) effort. The research conducted demonstrates that metadata has an inherent “predictive value” and confirms that metadata is crucial for data analytics. The work presented also provides insights into the best direction for future work in this area.