UMIACS Computational Linguistics Colloquium Series, Spring 2002

Time and Location

Talks are on Wednesdays at 10:00am unless otherwise noted.
The colloquium will be held in A.V. Williams Building, Room 2120 unless noted otherwise.
The schedule is always evolving. Titles and abstracts will be added as we receive them.

Date	Speaker	Affiliation	Title, abstract, other info
Special day/time/room: Jan 29, 3:45pm, AVW 3460	Bhuvana Ramabhadran	IBM	MALACH: Multilingual Access to Large spoken ArCHives
Feb 6	Rebecca Hwa, Philip Resnik, and Amy Weinberg	UMD	Evaluating Translational Correspondence using Annotation Projection
Feb 6, 2:30-3:30pm, AVW 3258 (extra talk, special time!)	Yuqing Gao	IBM	Statistical Approaches for Speech-to-Speech Translation
Feb 13, 9:30am (special time)	Lise Getoor	UMD	Learning Statistical Models from Relational Data
Feb 20	Fazil Ayan	UMD	Student practice talk: Generating a parsing lexicon
Feb 27	Nizar Habash	UMD	Student practice talk: Generation Heavy Machine Translation
March 6, 4pm (special time)	Lisa Pearl	UMD	Student practice talk: Improved Word-Level Alignment: Injecting Word-Level Knowledge about MT Divergences
Apr 10	Noah Smith	JHU	From Words to Corpora: Recognizing Translation
April 17	Carol Espy-Wilson	UMD	Knowledge-based Speech Recognition
April 24	Inderjeet Mani	MITRE/Georgetown	The Language of Time: Mining Text Corpora for Temporal Information
July 17	Mary Broman Olsen	Microsoft	Demonstration: German-English machine translation prototype built on Microsoft's Natural Language Software Development Kit
TBA	Jason Eisner	JHU	TBA (parsing)

For most speakers you can sign up for a meeting slot on the meetings sign-up page. If not, contact Denise Best (denise@cfar.umd.edu). or Philip Resnik (resnik@umiacs.umd.edu).

People who attend the CL colloquium series are encouraged to also attend the LAISEM (Logic and Artificial Intelligence) series Mondays at 11am, and the Linguistics colloquium series Fridays at 2pm; also check out the Linguistics Lunch Talks Thursdays at 12:30pm. Many talks in these series are likely to be of interest to computational linguists.

This series is sponsored by the University of Maryland Language and Media Processing Laboratory, under a contract from the Department of Defense.