Attribute selection for referring expression generation: new algorithms and evaluation methods

Albert Gatt, Anja Belz

Research output: Chapter in Book/Conference proceeding with ISSN or ISBNConference contribution with ISSN or ISBN

Abstract

Referring expression generation has recently been the subject of the first Shared Task Challenge in NLG. In this paper, we analyse the systems that participated in the Challenge in terms of their algorithmic properties, comparing new techniques to classic ones, based on results from a new human task-performance experiment and from the intrinsic measures that were used in the Challenge. We also consider the relationship between different evaluation methods, showing that extrinsic taskperformance experiments and intrinsic evaluation methods yield results that are not significantly correlated. We argue that this highlights the importance of including extrinsic evaluation methods in comparative NLG evaluations.
Original languageEnglish
Title of host publicationINLG '08 Proceedings of the Fifth International Natural Language Generation Conference
Place of PublicationStroudsburg, PA, USA
PublisherAssociation for Computational Linguistics
Pages50-58
Number of pages9
DOIs
Publication statusPublished - 1 Jan 2008
EventINLG '08 Proceedings of the Fifth International Natural Language Generation Conference - Salt Fork State Park, Ohio, USA
Duration: 1 Jan 2008 → …

Conference

ConferenceINLG '08 Proceedings of the Fifth International Natural Language Generation Conference
Period1/01/08 → …

Fingerprint Dive into the research topics of 'Attribute selection for referring expression generation: new algorithms and evaluation methods'. Together they form a unique fingerprint.

Cite this