Translation software that develops an understanding of languages by scanning through thousands of previously translated documents has been developed by Information Sciences Institute.

It takes a statistical approach, building probabilistic rules about words, phrases and syntactic structures.

The translated documents used to teach the translation algorithms can be electronic, on paper, or even audio files. The developer says the system is not only faster than other methods, but also better suited to tackling less common languages and the unusual vocabulary found in specialised or technical texts.