Jason R. Smith


Contact Info:


Email:
Office: Computational Science and Engineering Building 321


I'm a sixth-year Ph.D. student working under Chris Callison-Burch and Adam Lopez.  My thesis topic is on mining parallel data, and both of my internships (Microsoft Research in 2009 and Google in 2011) have been on this topic. My focus in this area is on low-resource languages. Previously I've worked with Jason Eisner on semi-supervised learning methods on large-scale data, and on morphology with Jason and Markus Dreyer. I've also worked on the GALE program with Damianos Karakos on system combination for machine translation.

Research Interests

As a masters student at Louisiana Tech University under my previous advisor, Andrei Paun, I was focused mostly on automata theory. Now I am interested in areas in natural language processing which include the application of finite state automata, such as the morphology work I've done with Markus Dreyer. My primary interest is in machine translation. For one year I've been on the DARPA GALE project, working in system combination for machine translation with Damianos Karakos. In 2009, I was an intern at Microsoft Research working with Chris Quirk and Kristina Toutanova, and in 2011 I interned at Google, where I worked with Peng Xu and Jakob Uszkoreit. With Jason Eisner, I worked on methods for large scale semi-supervised learning. In 2007, I was also a member of the ELERFED team in the CLSP Worksop 2007, and have since been interested in methods of extracting knowledge from Wikipedia.

Here is a current CV.



Publications

Moved to the new CLSP website