Acknowledgements for my Ph.D. dissertation

I have thoroughly enjoyed and benefited from my career as a graduate
student due to the strategic guidance of my advisor, Kathleen
R. McKeown.  Kathy encouraged me to explore various areas, and
ultimately provided priceless support in my chosen research topic.  I
am fortunate to have discovered the natural language problem that
interests me most, applying machine learning methods, my other primary
field of interest, to boot.

The other four members of my thesis committee also logged many hours
over the last few years, providing valuable insights and feedback for
my work.  Judith L. Klavans was my primary linguistic consultant and a
great source of enthusiasm.  Conversations about the linguistic side
of this work with Philip Resnik were also extremely stimulating.  The
two of them kept me on target by providing many linguistic
observations to further support my work, as well as teaching me the
standard linguistic protocols for testing and evaluation.  John
R. Kender, in general, served as a priceless guide through my six
years of graduate school teaching and research.  Moreover, John helped
me to more formally assess and defend semantic distinctions set forth
by linguists; the design of Section 2.2 resulted from discussions with
John.  Finally, Salvatore J. Stolfo was my primary machine learning
consultant.  He was like the Judith and Philip of learning; only due
to his insights is the machine learning side of this work meaningful
from the vantage of that field.

What can I say about Jacques Robin, my graduate officemate of 5 years?
Jacques became a close friend and a great basketball coach.  He showed me
how to balance the untamed ambitions of a new graduate student with the
pragmatics of actual execution, and he showed me how to balance a research
life with a night life.  Jacques is an inspired, hilarious spark of positive
action.  He gave me pep talks and companionship.  He's better than a pet
dog.

The remaining members of Columbia's natural language processing group
were incredibly important for feedback on this work.  Vasileios
Hatzivassiloglou was my primary statistical consultant; he is an
encyclopedia of mathematical models, and was a great source of
perspective on my work.  He helped formalize one interpretation of my
results, discussed in Section 2.2.4.  Also, Vasileios designed and
implemented the clustering of verbs described in Section 6.2.4.  James
Shaw first pointed out to me that "have" is statively ambiguous,
resulting in the work reported in Section 4.3.  Other helpful members
of Columbia's NLP group include or included Michael Elhadad, Eleazar
Eskin, Pascale Fung, Hongyan Jing, Min-Yen Kan, Diane Litman, Rebecca
Passennau, Dragomir Radev, Ruth Reeves, Nina Wacholder, and Robert
Weida.

Many other people in my life also lended a helping hand.  Three
professors from my undergraduate days at Brandeis University were
extremely formative in my development as a researcher, and invested a
lot of time to help me grow: Harry Mairson, Richard Alterman, and
James Pustejovsky.  Harry showed me the humor behind computer science
and helped me first understand advanced research material.  Rick was a
great undergraduate thesis advisor (language generation), and inspired
me to ride my bike to work every day.  James first got me interested
in linguistics, and taught me to indulge in the deep questions of
artificial intelligence.  Alexander Day Chaffee, my friend from
childhood days of hacking, helped me enormously with the
interpretation of numerical results, and even had a couple helpful
linguistic insights.  Alex knows what is interesting about life and
what is fun about computers; he will always be the Dalang of my
Gamelan.  Other important contributors include: Peter Angeline, Ken
Church, Ivan A. Derzhanski, Larry Eshelman, Jussi Karlgren, John Koza,
Nelson Minar, Andreas L. Prodromidis, Conor Ryan, Astro Teller, Andy
Singleton, David Schaffer, and Dakai Wu.

Thank you, my parents, Lisa A. Schamberg and Andrew Siegel, who are the
ultimate best you could want.  I appreciate their diligent struggle to find
the optimal frequency with which to ask, ``Is it done yet?''  My sister,
Rachel A. Siegel, is the center of the universe.  Also, I'd like to thank my
issues for remaining repressed (Alexander D. Chaffee, personal
communication).

This research is supported in part by the Columbia University Center for
Advanced Technology in High Performance Computing and Communications in
Healthcare (funded by the New York State Science and Technology
Foundation), the Office of Naval Research under contract N00014-95-1-0745
and by the National Science Foundation under contract GER-90-24069.