Data-Oriented Parsing and Generation

Hauptseminar: Computerlinguistik, 2. Studienabschnitt

Abstract

Data-Oriented Parsing (DOP) models embody the assumption that humans produce and interpret natural language utterances by invoking representations of their concrete past language experience, rather than the rules of a consistent and non-redundance competence grammar. DOP models therefore maintain large corpora of sentences with syntactic structures. They analyze new input-sentences by combining partial structures from the corpus, and employ the occurrence frequencies of these structures to estimate which of the resulting analysis are the most probable one. During this seminar we will have a closer look to the computational and linguistic aspects of DOP. Recently, first Data-Oriented methods for natural language generation have been proposed, which we will discuss at the end of the seminar.

Seminar language: English

Certificates
Referat(2 LP/4 LP) oder Referat + Hausarbeit (4 LP/9 LP) (talk or talk + paper)

Concerning the Master programme:
M.Sc. Program: Specialication Course L (linguistics)

Credit points:
Diplom/M.Sc.: talk only 2 LP/4 LP; talk + paper 4 LP/9 LP

Schedule (Topics)

Paper Report

• The paper lenght is restricted to eight (8) pages.
• Deadline for submission is the 30th of September 2005.
• Submission should be electronically made to my email address!

References

Session 01:

• Rens Bod and Remko Scha. A DOP model for phrase-structure trees. In Rens Bod, Remko Scha, and Khalil Sima'an, editors, Data-Oriented Parsing. CSLI Publications, Stanford, CA, to appear. [SchaDOP]
  
• Joshua Goodman “Semiring Parsing”, Computational Linguistics, 25/4, 1999.

Session 02:

• Khalil Sima'an. Computational complexity of disambiguation under DOP. In Rens Bod, Remko Scha, and Khalil Sima'an, editors, Data-Oriented Parsing. CSLI Publications, Stanford, CA, to appear. [SimaanDOP]

Session 03:

• Srinivas Bangalore. Localizing dependencies and supertagging. In Rens Bod, Remko Scha, and Khalil Sima'an, editors, Data-Oriented Parsing. CSLI Publications, Stanford, CA, to appear. [BangaloreDOP]

Session 04:

• David Chiang. Statistical parsing with an automatically extracted tree-adjoining grammar. In Rens Bod, Remko Scha, and Khalil Sima'an, editors, Data-Oriented Parsing. CSLI Publications, Stanford, CA, to appear. [ChiangDOP]

Session 05:

• Rens Bod and Ronald Kaplan. A DOP model for lexical-functional grammar representations. In Rens Bod, Remko Scha, and Khalil Sima'an, editors, Data-Oriented Parsing. CSLI Publications, Stanford, CA, to appear. [KaplanDOP]

Session 06:

• Guenter Neumann. A data-driven approach to head-driven phrase-structure grammar. In Rens Bod, Remko Scha, and Khalil Sima'an, editors, Data-Oriented Parsing. CSLI Publications, Stanford, CA, to appear. [NeumannDOP]
• Kiril Simov. HPSG-Based Annotation Scheme for Corpora Development and Parsing Evaluation. In: Proc. of the RANLP 2003 Conference, Borovets, Bulgaria, 10-12 September 2003. pages 432-439.

Session 07:

• Guenter Neumann: Applying Explanation-based Learning to Control and Speeding-up Natural Language Generation. In Proceedings of ACL/EACL-97, Madrid, 1997.
• Christer Samulesson: An Efficient Algorithm for Surface Generation. In Proceedings of 14th IJCAI, Montreal, Canada, 1995.

Session 08:

• Adwait Ratnaparkhi. (2000). Trainable Methods for Surface Natural Language Generation. Proceedings of the 1st Meeting of the North American Chapter of the Association of Computational Linguistics (NAACL 2000). Pages 194--201. Seattle, Washington, USA.
• Adwait Ratnaparkhi. (2001). Modeling informational novelty in a conversational system with a hybrid statistical and grammar-based approach to natural language generation. Proceedings of the NAACL Workshop on Adaptation in Dialogue Systems, June 4, 2001, Pittsburgh, PA, USA.

Links

Remko Scha's background material on Data-Oriented Parsing (DOP)

Probabilistic Grammars and Data-Oriented Parsing  (course by Detlef Prescher, Remko Scha, and Khalil Sima'an)

What is DOP? by Rens Bod