Discourse Data in DiET

The DiET project provides systematically constructed and annotated test items and associated tools, enabling fast system debugging and evaluation, and automatic linkage from test items to real corpora instances. This paper concentrates on the discourse test suite and its use. The discourse test suite covers discourse phenomena such as pronouns, def-inites and ellipsis. These can be used to evaluate the coverage and accuracy of implementations of anaphora resolution algorithms. We also examine the text prooling support within the Diet tools. Text Prooling identiies typical and salient corpus characteristics, e.g. the frequency and distribution of part of speech tags and vocabulary richness. Prooling also provides candidate sentences instantiating predeened syntactic phenomena. Prooling enables users to select test-items appropriate to their domain speciic corpus. The paper shows how the corpus search engine can be used to identify discourse phenomena in a corpus and presents concrete results of this evaluation scenario.

Keywords

NLP evaluation test items discourse

Domains

Linguistics

Complete list of metadata

Submission Type	File
Deposit type	Conference papers
Title	en Discourse Data in DiET
Abstract	en The DiET project provides systematically constructed and annotated test items and associated tools, enabling fast system debugging and evaluation, and automatic linkage from test items to real corpora instances. This paper concentrates on the discourse test suite and its use. The discourse test suite covers discourse phenomena such as pronouns, def-inites and ellipsis. These can be used to evaluate the coverage and accuracy of implementations of anaphora resolution algorithms. We also examine the text prooling support within the Diet tools. Text Prooling identiies typical and salient corpus characteristics, e.g. the frequency and distribution of part of speech tags and vocabulary richness. Prooling also provides candidate sentences instantiating predeened syntactic phenomena. Prooling enables users to select test-items appropriate to their domain speciic corpus. The paper shows how the corpus search engine can be used to identify discourse phenomena in a corpus and presents concrete results of this evaluation scenario.
Authors	Ian Lewin ¹ , Pierrette Bouillon ² , Sabine Lehmann ^{3, 2} , David Milward ¹ , Ludovic Tanguy ² 1 SRI - SRI International [Menlo Park] ( 87773 ) - 333 Ravenswood Avenue Menlo Park CA, 94025 - United States 2 ETI/TIM/ISSCO - Ecole de Traduction et d'Interprétation ( 100625 ) - 40, boulevard du Pont-d'Arve 1211 GENEVE 4 - SUISSE - Switzerland Université de Genève = University of Geneva ( 154620 ) 3 DFKI - Deutsches Forschungszentrum für Künstliche Intelligenz GmbH = German Research Center for Artificial Intelligence ( 258630 ) - Stuhlsatzenhausweg 3Campus D3_2D-66123 Saarbruecken - Germany
Fulltext language	English
Production date	1999
Science popularization	No
Proceedings	Yes
Peer-reviewed	Yes
Invited	No
Audience	International
Publication date	1999
Volume title	EACL Workshop on Linguistically Interpreted Corpora (LINC)
Conference or book title	EACL Workshop on Linguistically Interpreted Corpora (LINC)
Start conference date	1999
City	Bergen
Country	Norway
Domain	Humanities and Social Sciences/Linguistics
Keywords	en NLP, evaluation, test items, discourse

Main file

diet_discourse99.pdf ( 344.2 Ko )

Origin : Files produced by the author(s)

Ludovic Tanguy : Connect in order to contact the contributor

https://shs.hal.science/halshs-01322332

Submitted on: Friday, May 27, 2016 at 2:03:38 PM

Last modification on: Sunday, January 7, 2024 at 6:56:06 PM

Long-term archiving on: Sunday, August 28, 2016 at 10:13:54 AM

Dates and versions

halshs-01322332, version 1 (27-05-2016)

Identifiers

HAL Id : halshs-01322332 , version 1

Cite

Ian Lewin, Pierrette Bouillon, Sabine Lehmann, David Milward, Ludovic Tanguy. Discourse Data in DiET. EACL Workshop on Linguistically Interpreted Corpora (LINC), 1999, Bergen, Norway. ⟨halshs-01322332⟩