Chair of Language Technology

1

• Chair of Language Technology

2
Outline

• General information
• Staff
• Teaching
• Courses
• Supervision
• Research
• Fields
• Main results
• Participation in conferences
• Organizing of schools/workshops
• Future plans

3
General information

• Chair was opened 01.09.2001
• http//www.cs.ut.ee/koit/KT/
• Co-operation with the linguists of the University
  of Tartu since 1980s
• Informal research group of computational
  linguistics
• http//www.cl.ut.ee
• (head Professor of General Linguistics Haldur
  Õim)
• Research theme Computational and language
  resources for Estonian theoretical and
  applicational aspects
• Teaching courses of language technology/computatio
  nal linguistics

4
Staff

• Mare Koit (cand. math.) professor
• Tiit Roosmaa (cand. techn.) associate professor
• Heli Uibo, MSc lecturer
• Kaili Müürisep, PhD researcher

5
Teaching

• 33 courses MTAT.06.
• Some courses
• Language technology
• Artificial intelligence I, II
• Introduction to computational linguistics
• Mathematics for computational linguists I, II
• Heli computer handling, programming, object
  oriented programming (MTAT.03) computational
  morphology (FLEE.08)

6
Teaching (2)

• Courses of visiting lecturers (partly supported
  by Estonian Tiger University)
• 2003
• Tatjana Gavrilova (S. Petersburg). Introduction
  to Knowledge Engineering
• Graham Wilcock (Helsinki), Kristiina Jokinen
  (Helsinki). XML-based Natural Language Generation
  
• Vadim Stefanyuk (Moscow). Lisp, Symbolic
  Programming Language
• 2004
• Graham Wilcock.
• Vadim Stefanyuk. Lisp and Artificial Intelligence
  
• Martin Volk (Stockholm), Karin Harbusch
  (Koblenz). Empirical Methods in Natural Language
  Processing
• 2005
• Yorick Wilks (Sheffield). Human-Computer Dialogue
  Processing, February 21-25
• Norbert Fuchs (Zurich). Attempto Controlled
  English, November 7-11

7
Teaching (3)

• Graduate students can participate in the courses
  of Nordic Graduate School of Language Technology
  (NGSLT) http//www.ngslt.org/
• Agreement concerning collaboration with the NGSLT
  between NGSLT and University of Tartu was signed
  28.10.2004
• Estonian Graduate School of Linguistics and
  Language Technology (2005-2008)
• offers courses, supports mobility of graduate
  students and visiting lecturers

8
Teaching (4)

• Theses defended since 01.09.2001
• 1 PhD (Tiina Puolakainen, 2001)
• 3 MSc (Kaarel Kaljurand, Margus Treumuth, Andrei
  Ivanov)
• 1 MA (Liina Eskor FLEE)
• 16 BSc (4 years)
• 1 BA (4 years)
• 4 BSc (3 years)
• 2 diploma theses
• Under supervision
• 6 PhD theses (Heli Uibo, Kaarel Kaljurand, Margus
  Treumuth, Liina Eskor, Maret Valdisoo, Evely
  Vutt)
• 4 MSc 42 (Aigar Sirel, Aleksei Ivanov, Taavet
  Kikas, Eva Valk)
• 1 MA 42 (Helen Nigol - FLEE)
• 2 MSc 32 (Pilleriin Mutso, Krista Liin)
• 1 MA 32 (Siiri Pärkson FLEE)

9
Research

• Computational morphology of Estonian (Tiit
  Roosmaa, Heli Uibo)
• Computational syntax of Estonian (Tiit Roosmaa,
  Kaili Müürisep, Heli Uibo)
• 3) Computational semantics (PhD student Kaarel
  Kaljurand)
• 4) Dialogue modelling (Mare Koit)

10
Research (2)

• (1) Computational morphology
• Two-level morphology of Estonian
• (2) Computational syntax
• Surface syntactic analyser of written Estonian
• Syntactically annotated text corpus
• Adaptation of the surface syntactic analyser to
  spoken Estonian started
• Estonian treebank started
• Text summarizer which uses syntactic analysis
• Experiments of using of functional unification
  formalism for sentence generation

11
Research (3)

• (3) Computational semantics
• Software for word sense disambiguation
• Term extraction from Estonian texts using
  semantics and machine learning
• (4) Dialogue modelling
• Estonian dialogue corpus
• Software for processing of dialogue corpus
• Experiments on automatic recognition of dialogue
  acts using machine learning started
• Simple dialogue system implemented

12
Research (4)

• International projects
• Nordic Treebank Network 01.04.2003-30.06.2005
  (Heli, Kaili)
• PaNoLa-plus (Parsing Nordic Languages)
  01.09.2004-31.01.2005 (Heli)
• TELRI II 1999-2002 (Tiit)
• PhD student Kaarel Kaljurand is participating in
  2 international projects of the Zurich University
  
• Projects of Estonian Science Foundation
• 2 projects 2001-2003
• 1 project 2004 -gt (Modelling of conversational
  agent)
• National programme Estonian and national culture,
  language technology projects
• 2 projects 2002-2003
• National programme Estonian and national memory,
  language technology projects
• 2 projects 2004 -gt (Syntax-based language
  software and linguistic resources, Information
  dialogue with computer in Estonian)

13
Research (5)

• Participating in international conferences, e.g.
• 2004
• ACL-04 (Annual Meeting of the Association of
  Computatioal Linguistics) - Tiit
• LREC-04 (International Conference on Language
  Resources and Evaluation) - Tiit
• ECAI-04 (European Conference of Artificial
  Intelligence) - Mare
• COLING-04 (International Conference of
  Computational Linguistics) - Heli
• TLT 2004 (Treebanks and Linguistic Theories) -
  Heli
• 2005
• ACL-05 - Heli
• IJCAI-05 (International Conference of Artificial
  Intelligence) - Mare, Tiit
• NoDaLiDa-05 (Nordic Conference of Computational
  Linguistics) - Heli, Kaili, Mare
• RANLP-05 (Recent Advances in Natural Language
  Processing) - Heli

14
Organizing of schools/workshops

• 2002
• Summer school Language Technology in
  Human-Computer Interaction (19 participants 4
  from Russia)
• 2004
• Summer school Empirical Methods in Natural
  Language Processing (22 participants 4 from
  Latvia, 2 - Lithuania, 3 - Russia, 4 - Ukraine)
• Nordic Treebank Network workshop in Tartu (Heli)
• 2007 NoDaLiDa in Tartu

15
Future plans

• Teaching
• New courses
• 2005/06
• Theories and Models of Natural Language Syntax 4
  CP (Kaili, Heli)
• Syntactic Analyzer 2 CP (Kaili)
• Statistical Models of Natural Language 2 CP (Mare)

16
Future plans (2)

• Research
• Cooperation in University of Tartu (with
  department of Estonian and Finno-Ugric
  Linguistics)
• Cooperation in Estonia (with Tallinn University
  of Technology and Institute of Estonian Language)
• Graduate School of Linguistics and Language
  Technology
• National programme Computer support of Estonian
  (2006-2010)
• Machine translation
• Multilingual information retrieval
• Information dialogue with computer
• International cooperation
• NGSLT
• Nordic Treebank Network activities
• New projects?
• Multilingual text summarizing (Marie Curie
  Actions Research Training Networks)
• PaNoLa-Baltic software for teaching (Estonian)
  syntax (Research Council of Norway under the
  programme NORDPLUS neighbour)