GDELT represents both a living laboratory for studying the world in realtime and a platform for prototyping the future of technologies that allow us to better understand planetary-scale datasets in new ways in realtime.

As a living laboratory, GDELT provides a wealth of realtime open datasets spanning over 8 trillion datapoints across two billion online news articles spanning 20 years and 152 languages, half a billion worldwide news images, a decade of television news over portions of 150 channels and half a decade of radio news covering portions of 550 stations. This incredibly unique dataset today powers an ever-growing fraction of the world's realtime risk assessment and forecasting and is used for everything from forecasting forced human migration to sending one of the first alerts of Covid-19 to combatting Russian misinformation.

At the same time, GDELT's immense breadth, realtime updates, coverage of all modalities and massively multilingual scope make it the ultimate platform through which to prototype tomorrow's future technologies for content understanding. GDELT's datasets represent some of the largest real world non-synthetic open datasets spanning graphs, embeddings, AI annotations, geography and linguistics and update in realtime, making it possible to explore questions from realtime news recommender systems to optimized ANN and graph analytic algorithms to internationalizing NLP systems.

We'd love to hear how you're using GDELT and are happy to feature particularly interesting applications here on the GDELT blog!