2016/06/01 by Saurav Ghosh, Prithwish Chakraborty, Ghosh, Saurav +11
Medicine · Social Sciences · #Computation and Language (cs.CL) #Computational and Text Analysis Methods #Data-Driven Disease Surveillance #FOS: Computer and information sciences #Influenza Virus Research Studies #Information Retrieval (cs.IR) #Machine Learning (stat.ML) #Social and Information Networks (cs.SI)
paper · pdf · doi:10.48550/arxiv.1606.00411
openalex publication_date 2016/06/01 · openalex created_date 2022/10/04 · openalex updated_date 2026/07/28
In retrospective assessments, internet news reports have been shown to\ncapture early reports of unknown infectious disease transmission prior to\nofficial laboratory confirmation. In general, media interest and reporting\npeaks and wanes during the course of an outbreak. In this study, we quantify\nthe extent to which media interest during infectious disease outbreaks is\nindicative of trends of reported incidence. We introduce an approach that uses\nsupervised temporal topic models to transform large corpora of news articles\ninto temporal topic trends. The key advantages of this approach include,\napplicability to a wide range of diseases, and ability to capture disease\ndynamics - including seasonality, abrupt peaks and troughs. We evaluated the\nmethod using data from multiple infectious disease outbreaks reported in the\nUnited States of America (U.S.), China and India. We noted that temporal topic\ntrends extracted from disease-related news reports successfully captured the\ndynamics of multiple outbreaks such as whooping cough in U.S. (2012), dengue\noutbreaks in India (2013) and China (2014). Our observations also suggest that\nefficient modeling of temporal topic trends using time-series regression\ntechniques can estimate disease case counts with increased precision before\nofficial reports by health organizations.\n