Discrete mathematics, dynamic programming, combinatorial optimization, statistical natural language processing, formal languages and automata theory, combat zoology.
I am currently a member of the Random Disambiguation Paths
project in the AMS department at JHU.
The group studies the problem of finding shortest paths through a field of probabilistic hazards.
I've been working on a discrete version of the problem,
where the field is represented by a graph with some edges possibly unusable,
and the agent trying to traverse it is only allowed to check a limited number of edges.
In the past, I have worked on the Dyna project under the direction of
Jason Eisner.
Dyna is a simple, declarative language designed to make it easy to efficiently implement dynamic programming algorithms.
My research helped to show how to generalize algorithms discovered by NLP researchers for specific contexts as program transformations in Dyna's formalism.
I spent the summer of 2006 working in the research group at Google, where I was working on a summarization project related to the Google Book Search project. Details will hopefully be made available in a forthcoming paper.
In college, I was a member of the math department. My advisor there was Edward Nelson, under whose direction I wrote a thesis presenting "A New Proof of the Conservativeness of Internal Set Theory".
Non-research interests:
I play ultimate frisbee, a sport that the world ought to know more about,
something I've been doing since college. If you live in Baltimore, you should think about playing in the recreational leagues run by the Central Maryland Ultimate Association; they're friendly, fun, and a great way to get into the sport.
I also enjoy long-distance running, and am currently training for the 2009 Boston Marathon.
