Summer School in Tartu, 9.-14. August 2004

Institute of Computer Science, J. Liivi 2, Tartu
Lecture room: 206
Computer room: 205

Last modified 14.08.2004

Monday Tuesday Wednesday Thursday Friday Saturday

9:15 Opening

9:30 - 11:00 Karin Harbusch lecture: Motivation/Example domains for the exercises in the computer lab: Part-of-Speech (POS-) tagging and ambiguous typing with reduced keyboards
Basic definitions (random variables, ((un-)conditional probalilities, Bayes rules, MLE, ELE for context independent POS evaluation (test and trainings corpora, cross validation, accuracy ...)
Martin Volk lecture: Treebank Tools Karin Harbusch lecture: Revisiting the presented techniques and applying them for another application: Language models for ambiguous typing with reduced keyboards Excursion
to South-Estonia
Pictures Martin Volk lecture: The future of Treebanks Student presentations:
Erika Rimkute (PhD student)
An annotated Lithuanian corpus - morphological ambiguity and
disambiguation.
Gintare Grigonyte (MA student)
Verb phrase syntactical analyzer in Lithuanian language
Ansis Ataols Berzins (PhD student)
A creation of universal electronical-paper language-to-language dictionary

11:15 - 12:45 Martin Volk computerlab: Treebank Queries and Treebank Extraction Martin Volk computerlab: Finishing off ... Justine Boiko (BA student)
A Computer Program for an Automated Morphemic Morphological Analysis for
Latvian
Andris Liedskalnins (BA student)
Multilingual terminology portal
Marina Khoruzhenko & Daria Suslova (BA students)
German resources for generator QGen
Closing

12:45 - 13:45 lunch (Cafe Heller, Lossi St. 15) lunch lunch lunch lunch

13:45 - 16:15 Martin Volk lecture: An Introduction to Treebanking (incl. Corpus Collection and Preparation) Karin Harbusch lecture: Context-dependent PoS (n-grams, Markov chains, HMM) Martin Volk lecture: Treebank Usage (Training a Parser on a Treebank) Karin Harbusch Computer Lab: devolopping a modul for one of the following tasks:
- modifications for a given POS Tagger + evaluation
- modifications for a given ambiguous typing device + evaluation
- combining POS Tagging + n-gram model for ambiguous
typing + evaluation

16:30 - 18:00 Martin Volk computerlab: Annotating a Treebank Martin Volk computerlab: Evaluating a Chunker

evening 18:00
Welcome party
in the Botanical garden of the University of Tartu
(Lai St. 40) 19:00 dinner for participants from abroad (cafe Heller) 19:00 dinner for participants from abroad 19:00 dinner

Independent work in the computer room 205: Tuesday and Wednesday, 18:00-20:00

	Monday	Tuesday	Wednesday	Thursday	Friday	Saturday
9:15	Opening
9:30 - 11:00	Karin Harbusch lecture: Motivation/Example domains for the exercises in the computer lab: Part-of-Speech (POS-) tagging and ambiguous typing with reduced keyboards Basic definitions (random variables, ((un-)conditional probalilities, Bayes rules, MLE, ELE for context independent POS evaluation (test and trainings corpora, cross validation, accuracy ...)	Martin Volk lecture: Treebank Tools	Karin Harbusch lecture: Revisiting the presented techniques and applying them for another application: Language models for ambiguous typing with reduced keyboards	Excursion to South-Estonia Pictures	Martin Volk lecture: The future of Treebanks	Student presentations: Erika Rimkute (PhD student) An annotated Lithuanian corpus - morphological ambiguity and disambiguation. Gintare Grigonyte (MA student) Verb phrase syntactical analyzer in Lithuanian language Ansis Ataols Berzins (PhD student) A creation of universal electronical-paper language-to-language dictionary
11:15 - 12:45		Martin Volk computerlab: Treebank Queries and Treebank Extraction			Martin Volk computerlab: Finishing off ...	Justine Boiko (BA student) A Computer Program for an Automated Morphemic Morphological Analysis for Latvian Andris Liedskalnins (BA student) Multilingual terminology portal Marina Khoruzhenko & Daria Suslova (BA students) German resources for generator QGen Closing
12:45 - 13:45	lunch (Cafe Heller, Lossi St. 15)	lunch	lunch		lunch	lunch
13:45 - 16:15	Martin Volk lecture: An Introduction to Treebanking (incl. Corpus Collection and Preparation)	Karin Harbusch lecture: Context-dependent PoS (n-grams, Markov chains, HMM)	Martin Volk lecture: Treebank Usage (Training a Parser on a Treebank)		Karin Harbusch Computer Lab: devolopping a modul for one of the following tasks: - modifications for a given POS Tagger + evaluation - modifications for a given ambiguous typing device + evaluation - combining POS Tagging + n-gram model for ambiguous typing + evaluation
16:30 - 18:00	Martin Volk computerlab: Annotating a Treebank		Martin Volk computerlab: Evaluating a Chunker
evening	18:00 Welcome party in the Botanical garden of the University of Tartu (Lai St. 40)	19:00 dinner for participants from abroad (cafe Heller)	19:00 dinner for participants from abroad		19:00 dinner