My Research

Jason Michelizzi
Department of Computer Science
University of Minnesota Duluth
320 Heller Hall
1114 Kirby Drive
Duluth, MN  55812-2496

I am a graduate student in Computer Science at the University of Minnesota Duluth and a Research Assistant for Dr. Ted Pedersen. My research is in the field of Natural Language Processing, and I am a member of the NLP Group at UMD.

My thesis proposal: PostScript or PDF (Warning: the converter used to convert from PS to PDF sometimes generates low-quality PDFs)

Software I've developed

WordNet::Similarity: a Perl module that computes semantic relatedness
A web interface for WordNet::Similarity
SenseRelate: Perl software for Word Sense Disambiguation
Text-Similarity: a Perl module for measuring the similarity of text documents. This now seems to be the primary focus of my thesis research.
A patch for WordNet::QueryData to add a frequency function that gets the frequency of a synset out of WordNet's cntlist. Update: as of version 1.37, my modification comes with the "standard" version of WordNet::QueryData, so you no longer need to apply the patch manually.
randomPairs.pl: a Perl program that generates a list of pairs of randomly select word senses from WordNet
Tools for UMLS®
wn.pl: a command-line interface to WordNet (not terribly useful unless you have trouble with the wn program that comes with WordNet)

Perl Monks
PerlGuts Illustrated
A Perl Monks thread on doing OOP in Perl the right way
The defined-or patch for recent versions of Perl (be sure to get the right patch for your version of Perl; the most recent is dor-5.8.6.diff). The defined-or operator is very useful and will soon become part of "standard" Perl. Okay, technically it's three operators: //, //=, and err. Read more about why the defined-or operator is cool.