In science fiction movies people can always talk with computers (and aliens speaking different languages). What kinds of things can one do in reality? Recent data-driven probabilistic approaches to natural language processing (NLP) have led to programs that can do many parts of the problem very well. It is still hard to completely understand human languages, but why should we settle for doing nothing at all? In this talk I particularly want to develop the idea that now that we have gigahertz of processing power and hundreds of gig of disk, there are lots of places where "systems programs" could use a little embedded NLP. The 'file' command could do with some NLP. Ghostscript (ps2ascii) could definitely do with some NLP. Getting control of one's email back really needs NLP.
About the speaker:
Christopher Manning is an assistant professor of computer science and linguistics at Stanford University. Previously, he has held faculty positions at Carnegie Mellon University and the University of Sydney. His research interests include probabilistic natural language processing, syntax, information extraction, and computational lexicography. He is the author of three books, including Foundations of Statistical Natural Language Processing (MIT Press, 1999, with Hinrich Schütze).
Dept of Computer Science, Gates 4A
Stanford CA 94305-9040