A novel use of MT in the development of a text level analytic for language learning

In the current budget climate, it is important to investigate multiple uses for technologies that have benefitted from U.S. Government investment. Authentic materials indexed at appropriate learning levels are a requirement of several foreign language training activities. In this paper, we propose an approach that automatically annotates texts with language proficiency/ difficulty levels. Our approach is novel in that it uses independently available machine translation output of the source language into English coupled with an English-trained automatic text leveling analytic. This approach precludes text leveling annotation for each new language, though it requires a machine translation system. We report our initial results for Farsi document leveling. This automatic system is introduced as part of a wider adaptive platform currently under development called LanguageNation.