I am a PhD student in the Natural Language Processing (NLP) group, advised by Professor Kathleen McKeown. My research lies at the intersection of NLP, machine translation (MT) and information retrieval.

Specifically, I am interested in task-oriented machine translation: problems that require translated answers, such as cross-lingual IR, cross-lingual semantic role labeling, or cross-lingual question-answering. How can we evaluate tasks that require translation? Can the context of the task be used to improve or post-edit the MT? Conversely, by using bilingual information from the source data and machine translated data, can we improve performance on a specific task?

In addition to my research on task-oriented MT, I have also worked on real-life, large-scale applications of machine translation at Yahoo! and Google. Even though MT is often imperfect, it can allow users to access information that would otherwise be totally unreadable or unavailable. In summer 2009, my project at Google involved an indexing architecture for cross-lingual web search. The previous summer, I worked on a translation feature for maps.google.com. At Yahoo in 2006, I worked on cross-lingual image search.