Spatio-temporal disease tracking using news articles

Geographical Information Systems have been increasingly used to aid the prompt detection, tracking, and analysis of disease outbreaks. Web content which is full of health-related data also serves as a useful resource for disease outbreak analysis. News posts often report the initial outbreak of diseases and contain valuable information that aids in ascertaining the time and location of the disease outbreak. The locations mentioned in the news posts are specified textually rather than geometrically thereby requiring the use of geotagging methods to detect them and to map the textual specification to the corresponding actual geometric specification. The NewsStand system which aggregates news posts by topic and location while providing a map query interface to them is enhanced to enable disease tracking and analysis by geotagging disease-related web news posts. Besides the powerful functionalities of NewsStand for news exploration, enhancements of NewsStand with respect to the analysis of temporal information are described which include a well-designed time slider, a heatmap-based visualization tool for displaying disease distribution, and intuitive spatio-temporal querying methods. Future improvements to NewsStand are also discussed.