A corpus-based methodology for evaluating metrics of coherence for text structuring

Nikiforos Karamanis, Chris Mellish, Jon Oberlander, Massimo Poesio

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

This paper presents a novel corpus-based methodology for comparing metrics of coherence with respect to their potential usefulness for text structuring. Different definitions of such metrics, all using notions from Centering Theory, are discussed, and the methodology is applied to existing data derived from the MPIRO generation system.

Original languageEnglish
Title of host publicationNatural Language Generation
Subtitle of host publicationThird International Conference, INLG 2004, Brockenhurst, UK, July 14-16, 2004. Proceedings
EditorsAnja Belz, Roger Evans, Paul Piwek
PublisherSpringer
Pages90-99
Number of pages10
ISBN (Electronic)9783540278238
ISBN (Print)9783540223405
DOIs
Publication statusPublished - 2004
EventThird International Conference on Natural Language Generation, INLG 2004 - Brockenhurst, United Kingdom
Duration: 14 Jul 200416 Jul 2004

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume3123
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceThird International Conference on Natural Language Generation, INLG 2004
CountryUnited Kingdom
CityBrockenhurst
Period14/07/0416/07/04

Cite this

Karamanis, N., Mellish, C., Oberlander, J., & Poesio, M. (2004). A corpus-based methodology for evaluating metrics of coherence for text structuring. In A. Belz, R. Evans, & P. Piwek (Eds.), Natural Language Generation: Third International Conference, INLG 2004, Brockenhurst, UK, July 14-16, 2004. Proceedings (pp. 90-99). (Lecture Notes in Computer Science; Vol. 3123). Springer . https://doi.org/10.1007/978-3-540-27823-8_10