An Incremental Syntactic Language Model for Statistical Phrase-based Machine Translation

During my time as a PhD student at the University of Minnesota, I often told others that I was the only full-time machine researcher between Pittsburgh and Stanford. While I eventually learned of a handful of exceptions to that statement (most notably in Dayton, Ohio, where I am now employed), in retrospect my assumption that I could successfully pursue research in machine translation at an institution with no machine translation faculty seems remarkably naive. My eventual success in this achievement is due in no small part to my advisor, William Schuler, who supported and encouraged me in my research. It was on his suggestion that I set out in 2005 to implement the first open source implementation of hierarchical phrase-based machine translation, based solely on the description in the original paper (Chiang, 2005). So it was that I became perhaps the only machine translation researcher ever to be introduced to statistical machine translation through hierarchical phrase-based models, and only later become aware of the more straightforward word-based and phrase-based models. In the spring of 2008, I learned of an upcoming event — the 2nd Machine Translation Marathon, to be held in May in Wandlitz, Germany. I owe my advisor a tremendous debt of gratitude for providing funds for me to attend. My decision to attend this event was the single most important decision of m