From 1994 through 1997, I was an Associate Scientist at the Speech and Language Processing group at BBN in Cambridge, Massachusetts. At BBN, I helped create an architecture for porting NLP applications from one domain to another, and I co-created and patented the first state-of-the-art stochastic named entity finder, Nymble (now known as IdentiFinder(tm)).
While pursuing my doctorate at the University of Pennsylvania, I built the first extensible syntactic parsing engine, discovering new and surprising properties of syntactic parsing models.
From October, 2004 to March, 2010, I was a Research Staff Member (a “research scientist”) at the IBM T. J. Watson Research Center. My job was to discover new methods for the processing of human language text, a.k.a., “natural language processing”, or NLP for short. The methods I typically explored were statistical.
As of May, 2010, I have been a Research Scientist at Google, continuing to do work in statistical NLP.