Probabilistic generation of weather forecast texts

Anja Belz

Probabilistic generation of weather forecast texts

Anja Belz

University of Brighton

Research output: Chapter in Book/Conference proceeding with ISSN or ISBN › Conference contribution with ISSN or ISBN › peer-review

Abstract

This paper reports experiments in which pC RU — a generation framework that combines probabilistic generation methodology with a comprehensive model of the generation space — is used to semi-automatically create several versions of a weather forecast text generator. The generators are evaluated in terms of output quality, development time and computational efficiency against (i) human forecasters, (ii) a traditional handcrafted pipelined NLG system, and (iii) a HALOGEN-style statistical generator. The most striking result is that despite acquiring all decision-making abilities automatically, the best pC RU generators receive higher scores from human judges than forecasts written by experts.

Original language	English
Title of host publication	Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics
Place of Publication	Stroudsburg, PA, USA
Publisher	Association for Computational Linguistics
Pages	164-171
Number of pages	8
Publication status	Published - 1 Jan 2007
Event	Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics - Rochester, New York, USA Duration: 1 Jan 2007 → …

Conference

Conference	Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics
Period	1/01/07 → …

Bibliographical note

Creative Commons Attribution-NonCommercial-ShareAlike 3.0 License . Permission is granted to make copies for the purposes of teaching and research.

Access to Document

belz-hltnaacl07.pdfOther version, 76.6 KBLicence: CC BY

http://www.aclweb.org/anthology-new/N/N07/N07-1021.pdfLicence: Unspecified

Cite this

@inproceedings{3e877352aae54b49bc8a4d36bd986965,

title = "Probabilistic generation of weather forecast texts",

abstract = "This paper reports experiments in which pC RU — a generation framework that combines probabilistic generation methodology with a comprehensive model of the generation space — is used to semi-automatically create several versions of a weather forecast text generator. The generators are evaluated in terms of output quality, development time and computational efficiency against (i) human forecasters, (ii) a traditional handcrafted pipelined NLG system, and (iii) a HALOGEN-style statistical generator. The most striking result is that despite acquiring all decision-making abilities automatically, the best pC RU generators receive higher scores from human judges than forecasts written by experts.",

author = "Anja Belz",

note = "Creative Commons Attribution-NonCommercial-ShareAlike 3.0 License . Permission is granted to make copies for the purposes of teaching and research.; Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics ; Conference date: 01-01-2007",

year = "2007",

month = jan,

day = "1",

language = "English",

pages = "164--171",

booktitle = "Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics",

publisher = "Association for Computational Linguistics",

}

Belz, A 2007, Probabilistic generation of weather forecast texts. in Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics. Association for Computational Linguistics, Stroudsburg, PA, USA, pp. 164-171, Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics, 1/01/07. <http://www.aclweb.org/anthology-new/N/N07/N07-1021.pdf>

Probabilistic generation of weather forecast texts. / Belz, Anja.
Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics. Stroudsburg, PA, USA: Association for Computational Linguistics, 2007. p. 164-171.

Research output: Chapter in Book/Conference proceeding with ISSN or ISBN › Conference contribution with ISSN or ISBN › peer-review

TY - GEN

T1 - Probabilistic generation of weather forecast texts

AU - Belz, Anja

N1 - Creative Commons Attribution-NonCommercial-ShareAlike 3.0 License . Permission is granted to make copies for the purposes of teaching and research.

PY - 2007/1/1

Y1 - 2007/1/1

N2 - This paper reports experiments in which pC RU — a generation framework that combines probabilistic generation methodology with a comprehensive model of the generation space — is used to semi-automatically create several versions of a weather forecast text generator. The generators are evaluated in terms of output quality, development time and computational efficiency against (i) human forecasters, (ii) a traditional handcrafted pipelined NLG system, and (iii) a HALOGEN-style statistical generator. The most striking result is that despite acquiring all decision-making abilities automatically, the best pC RU generators receive higher scores from human judges than forecasts written by experts.

AB - This paper reports experiments in which pC RU — a generation framework that combines probabilistic generation methodology with a comprehensive model of the generation space — is used to semi-automatically create several versions of a weather forecast text generator. The generators are evaluated in terms of output quality, development time and computational efficiency against (i) human forecasters, (ii) a traditional handcrafted pipelined NLG system, and (iii) a HALOGEN-style statistical generator. The most striking result is that despite acquiring all decision-making abilities automatically, the best pC RU generators receive higher scores from human judges than forecasts written by experts.

M3 - Conference contribution with ISSN or ISBN

SP - 164

EP - 171

BT - Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics

PB - Association for Computational Linguistics

CY - Stroudsburg, PA, USA

T2 - Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics

Y2 - 1 January 2007

ER -

Probabilistic generation of weather forecast texts

Abstract

Conference

Bibliographical note

Access to Document

Fingerprint

Cite this