PowerPoint Presentation

GDELT: A Global Catalog of Human Society

Kalev LeetaruYahoo! Fellow in ResidenceGeorgetown Universityhttp://www.kalevleetaru.com/GDELT Team: Kalev Leetaru (Georgetown), Philip Schrodt (Parus Analytical Systems), Patrick Brandt (UTD)

1/3 global population online

As many cell phones as people on earth

Facebook alone has:

240 billion photographs (35% of all online photos)

1 billion members with 1 trillion connections

Our Digital WorldFast forward to today and weve reached an incredible point in human history.86.1 trillion text messages

2.2 trillion cell minutes in the US alone

107 trillion emails

1.6 million days worth of video uploaded to YouTube

Every Year2.5 billion new items added to Facebook

300 million photos posted to Facebook

500TB of new data about societys innermost thoughts posted to Facebook

As many words posted to Twitter every day as the entire New York Times in the last half-century

100 billion+ social media actions taken

Every Day600 new websites created

204 million emails sent

700,000 shares on Facebook

200,000 photos posted to Facebook

277,000 tweets sent

Every Minute

The Rise of Web News

GDELT:A Realtime Social Sciences Earth ObservatoryGoal: process all media worldwide into a single global catalog of behavior and beliefs in realtimeTeam: Kalev Leetaru (Georgetown), Philip Schrodt (Parus Analytical Systems), Patrick Brandt (UTD)What is GDELT?Physical Event Catalog: quarter-billion events in 300+ categories for all countries 1979-presentEmotional/Thematic Catalog: georeferenced catalog of thousands of emotions and themes worldwide over the same period to contextualize events (food riot vs political protest) and global reactionWhat is GDELT?Monitor all available news media worldwide (experimenting with citizen and social media)Translate 79 languages in realtime into English (Google Translate for Research award)Identify every mention of 300+ categories of events worldwide into a massive quarter-billion-record catalog of human societyIdentify latent narrative dimensions (emotion and thematic undercurrents) (GDELT 2.0)Biographical profiles (GDELT 2.0)Network dynamics (GDELT 2.0)

GDELTBehavior + BeliefsWhat is GDELT?Collection of quarter-billion georeferenced dyadic event records of X did Y to Z across the world in 300+ category CAMEO formatMaterial Conflict, Material Cooperation, Verbal Conflict, Verbal CooperationProvides physical behavior signal to study against beliefs/narrative media signalsEssentially takes a news report about a riot and puts it on a map with all of the associated detailsTwenty former soldiers were arrested for violent anti-government protests in front of the courthouse in Abuja, Nigeria yesterday. => Violent Protest+Arrest, Abuja Courthouse (Nigeria), 20 SoldiersOpen academic equivalent to US DOD ICEWSSourcesAll international news coverage from AfricaNews, Agence France Presse, Associated Press Online, Associated Press Worldstream, Associated Press, Christian Science Monitor, Facts on File, Foreign Broadcast Information Service, New York Times, United Press International, and the Washington Post.BBC Monitoring - global broadcast+print translated material.Google News - live realtime stream of all major international, national, regional, and local web-accessible news media.Google Translate for Research award to translate global news in realtime.Coversall countriesgloballyCovers a quarter-century:1979 to presentDaily updatesevery day, 365 days a yearBased on cross-section ofall major international, national, regional, local, and hyper-local news sources, bothprint and broadcast, from nearlyevery corner of the globe, in bothEnglish and vernacular58 fieldscaptureall available detailabout event and actorsTen fields capture significantdetail about each actor, includingroleandtypeAll recordsgeoreferenced to the city or landmarkas recorded in the articleSophisticatedgeographic pipeline disambiguatesandaffiliatesgeography with actorsSeparate geographic information forlocation of eventand forboth actors, includingGNSandGNISidentifiersAll records includeethnic and religious affiliationof both actors as provided in the textEven capturesambiguous events in conflict zones("unidentified gunmen stormed the mosque and killed 20 civilians")Specializedfilteringandlinguistic rewritingfilters considerablyenhance TABARI's accuracyWide array of media and emotion-based "importance" indicators for each eventNearly aquarter-billion eventrecords100% open, unclassified, and available for unlimited use and redistribution24

John Beieler Global 2013 Protests

GDELT 1979-2013 vs NASA Night LightsRolf Fredheim Russia Network

Jay Yonamine Afghanistan

New Scientist Magazine - Syria

John Beieler Egypt 2013

John Beieler 1979-2013 Protests

Interactive Data-Scale ExplorationCurrently piloting Googles Big Query database service with complete GDELT database.Live interactive complex queries across all quarter-billion records return in just 1-6s.Explore global patterns across quarter-century in realtime.

GDELTBehavior + BeliefsNot just what happened but how did people feel about itEgyptHosni Mubarak

Kalev Leetaru

Yahoo! Fellow in ResidenceGeorgetown University

Kalev.leetaru5@gmail.com

http://www.kalevleetaru.com/Event Data for Strategic ForecastingGDELT Team: Kalev Leetaru (Georgetown), Philip Schrodt (Parus Analytical Systems), Patrick Brandt (UTD)