27/01/2017

U.S. Intelligence Seeking Universal Text Translator

The U.S. intelligence community’s technology research arm, IARPA, is seeking bids for a program to develop a universal text translation search tool as part of their Machine Translation for English Retrieval of Information in Any Language (MATERIAL) program. The goal is to set up an intelligence search engine where analysts can input keywords in English, and the system will search available intelligence in any of the more than 7,000 languages worldwide, giving translated output in English. The intelligence retrieval program would sort documents based on a two-part query: first providing the domain (i.e. “government”) and the second a phrase descriptive of the specific information sought, like keywords such as “Zika virus.”
The Cipher Take: The move by IARPA is a step toward addressing the issue of so-called “low resource” languages, or languages where the intelligence community does not have significant numbers of speakers, so that when operating in remote areas, analysts are able to quickly search and translate documents. The push for a searchable universal translator with English in and English out could have lasting impact for intelligence analysis, but the road there is full of the human nuances inherent in over 7,000 languages worldwide—quite a tall order.

https://www.thecipherbrief.com/subscribe

ΣΧΟΛΙΟ "ΙΣΧΥΣ": Η ΕΛΛΗΝΙΚΗ (ΣΕ ΟΛΕΣ ΤΙΣ ΜΟΡΦΕΣ ΤΗΣ), ΚΑΛΥΠΤΕΤΑΙ ΕΠΑΡΚΩΣ ΑΠΟ ΤΑ ΗΔΗ ΥΠΑΡΧΟΝΤΑ ΚΑΙ ΛΕΙΤΟΥΡΓΟΥΝΤΑ ΜΕΤΑΦΡΑΣΤΙΚΑ "ΜΗΧΑΝΑΚΙΑ". ΑΥΤΟ ΓΙΑ ΣΥΖΗΤΗΣΗ ΠΟΥ ΕΙΧΑΜΕ ΤΙΣ ΠΡΟΑΛΛΕΣ ΜΕ ΚΑΠΟΙΟΥΣ ΠΟΥ ΑΙΣΘΑΝΟΝΤΑΝ ΑΣΦΑΛΕΙΣ ΕΠΕΙΔΗ ΧΡΗΣΙΜΟΠΟΙΟΥΣΑΝ ΑΡΧΑΙΑ ΕΛΛΗΝΙΚΑ ΣΤΙΣ ΜΕΤΑΞΥ ΤΟΥΣ ΕΠΙΚΟΙΝΩΝΙΕΣ. (ΚΑΙ ΤΑ ΛΑΤΙΝΙΚΑ). ΝΑ ΤΟ ΕΧΕΤΕ ΥΠΟΨΙΝ.

No comments :

Post a Comment