Evaluation sets for the MT track
The IWSLT 2013 Evaluation Campaign includes the MT trackon TED Talks. In this edition, the official language pairs are three:
from English to French
from German to English
from English to German
while twelve optional pairs in both directions are proposed:
English to/from Arabic, Chinese, Dutch, Italian, Persian, Polish, Portuguese (Brazilian), Romanian, Russian, Slovenian, Spanish, Turkish
Submitted runs on optional pairs will be evaluated as well, in the hope to stimulate the MT community to evaluate systems on common benchmarks and to share achievements on challenging translation tasks.
The archive with test sets is available at this link.
If you use this corpus in your work, please cite the paper:
M. Cettolo, C. Girardi, and M. Federico. 2012. WIT3: Web Inventory of Transcribed and Translated Talks. In Proc. of EAMT, pp. 261-268, Trento, Italy. pdf, bib.