TU Darmstadt / ULB / TUbiblio

Counting What Counts: Decompounding for Keyphrase Extraction

Erbs, Nicolai and Santos, Pedro Bispo and Zesch, Torsten and Gurevych, Iryna (2015):
Counting What Counts: Decompounding for Keyphrase Extraction.
In: Proceedings of the ACL 2015 Workshop on Novel Computational Approaches to Keyphrase Extraction, Association for Computational Linguistics, Beijing, China, ISBN 978-1-941643-62-4,
[Online-Edition: http://www.aclweb.org/anthology/W15-3603],
[Conference or Workshop Item]

Abstract

A core assumption of keyphrase extraction is that a concept is more important if it is mentioned more often in a document. Especially in languages like German that form large noun compounds, frequency counts might be misleading as concepts “hidden” in compounds are not counted. We hypothesize that using decompounding before counting term frequencies may lead to better keyphrase extraction. We identified two effects of decompounding: (i) enhanced frequency counts, and (ii) more keyphrase candidates. We created two German evaluation datasets to test our hypothesis and analyzed the effect of additional decompounding for keyphrase extraction.

Item Type: Conference or Workshop Item
Erschienen: 2015
Creators: Erbs, Nicolai and Santos, Pedro Bispo and Zesch, Torsten and Gurevych, Iryna
Title: Counting What Counts: Decompounding for Keyphrase Extraction
Language: English
Abstract:

A core assumption of keyphrase extraction is that a concept is more important if it is mentioned more often in a document. Especially in languages like German that form large noun compounds, frequency counts might be misleading as concepts “hidden” in compounds are not counted. We hypothesize that using decompounding before counting term frequencies may lead to better keyphrase extraction. We identified two effects of decompounding: (i) enhanced frequency counts, and (ii) more keyphrase candidates. We created two German evaluation datasets to test our hypothesis and analyzed the effect of additional decompounding for keyphrase extraction.

Title of Book: Proceedings of the ACL 2015 Workshop on Novel Computational Approaches to Keyphrase Extraction
Publisher: Association for Computational Linguistics
ISBN: 978-1-941643-62-4
Uncontrolled Keywords: UKP_p_WIKULU;UKP_p_DKPro;UKP_a_NLP4Wikis;UKP_reviewed;UKP_s_DKPro_Core
Divisions: 20 Department of Computer Science
20 Department of Computer Science > Ubiquitous Knowledge Processing
Event Location: Beijing, China
Date Deposited: 31 Dec 2016 14:29
Official URL: http://www.aclweb.org/anthology/W15-3603
Identification Number: TUD-CS-2015-0127
Export:
Suche nach Titel in: TUfind oder in Google

Optionen (nur für Redakteure)

View Item View Item