2014/02/27 by Naren Ramakrishnan, Ramakrishnan, Naren, P. J. Butler +59 · 1 citation
Computer Science · Medicine · #Anomaly Detection Techniques and Applications #Computers and Society (cs.CY) #Data Analysis with R #Data-Driven Disease Surveillance #FOS: Computer and information sciences #FOS: Physical sciences #I.2.7 #J.4 #K.4.1 #Physics and Society (physics.soc-ph) #Social and Information Networks (cs.SI)
paper · pdf · doi:10.48550/arxiv.1402.7035
openalex publication_date 2014/02/27 · openalex created_date 2022/08/27 · openalex updated_date 2026/07/28
We describe the design, implementation, and evaluation of EMBERS, an\nautomated, 24x7 continuous system for forecasting civil unrest across 10\ncountries of Latin America using open source indicators such as tweets, news\nsources, blogs, economic indicators, and other data sources. Unlike\nretrospective studies, EMBERS has been making forecasts into the future since\nNov 2012 which have been (and continue to be) evaluated by an independent T&E\nteam (MITRE). Of note, EMBERS has successfully forecast the uptick and downtick\nof incidents during the June 2013 protests in Brazil. We outline the system\narchitecture of EMBERS, individual models that leverage specific data sources,\nand a fusion and suppression engine that supports trading off specific\nevaluation criteria. EMBERS also provides an audit trail interface that enables\nthe investigation of why specific predictions were made along with the data\nutilized for forecasting. Through numerous evaluations, we demonstrate the\nsuperiority of EMBERS over baserate methods and its capability to forecast\nsignificant societal happenings.\n