Machine Translation is an old research topic that has recently attracted new interest and ideas. Statistical approaches in particular have evolved rapidly in the last few years, to the point where they now predominate at the yearly MT evaluations run by the U.S. National Institute of Standards and Technology (NIST), a focal point for MT research.

In this talk I will describe Portage, a state-of-the-art statistical MT system developed at NRC. Key techniques used by Portage include phrase-based translation models, dynamic-programming search, log-linear model combination, and error-driven rescoring. I will describe the system's recent participation in a dry-run NIST Chinese-to-English evaluation and give examples of its performance. The talk will conclude with an overview of ongoing MT research at NRC.