We are tremendously excited to announce that the long-awaited Translingual 2.0, with its new in-house language detector that supports 440 language-script combinations and its massive new 109-language and dialect machine translation architecture, with its powerful new many-to-many translation graph, new approach to nounphrase recovery, emphasis on rare words and geographic recovery, "minimal meaning unit" approach to scriptio continua languages and reach into a wealth of new languages is now online!
This immense new infrastructure will unlock an incredible array of new capabilities, especially when coupled with the new Web News NGrams 3.0 Dataset and usher in a fundamentally new era in global horizon scanning.
The full Translingual 2.0 infrastructure is now online as of today and running in production. As we complete the final stages of GDELT 3.0's launch this spring, we will be moving each GDELT dataset over to this new infrastructure module by module, so stay tuned!