Joint Arabic segmentation and part-of-speech tagging

John McNaught; Shabib AlGahtani

Joint Arabic segmentation and part-of-speech tagging

Ministry of Interior

Research output: Chapter in Book/Conference proceeding › Conference contribution › peer-review

Abstract

Arabic has a very complex morphological system, though a very structured one. Character patterns are often indicative of word class and word segmentation. In this paper, we explore a novel approach to Arabic word segmentation and part-of-speech tagging relying on character information. The approach is lexicon-free and does not require any morphological analysis, eliminating the factor of dictionary coverage. Using character-based analysis, the developed system yielded state-of-the-art accuracy comparing favourably with other taggers that involve external re-sources.

Original language	English
Title of host publication	Proceedings of the Second Workshop on Arabic Natural Language Processing
Editors	Nizar Habash, Stephan Vogel, Kareem Darwish
Place of Publication	Stroudsburg, USA
Publisher	Association for Computational Linguistics
Pages	108-117
Number of pages	10
ISBN (Print)	978-1-941643-58-7
Publication status	Published - 30 Jul 2015
Event	ACL Second Workshop on Arabic Natural Language Processing (WANLP 2015) - Beijing, China Duration: 30 Jul 2015 → 30 Jul 2015 http://https://aclweb.org/anthology/W/W15/W15-3212.pdf

Conference

Conference	ACL Second Workshop on Arabic Natural Language Processing (WANLP 2015)
City	Beijing, China
Period	30/07/15 → 30/07/15
Internet address	http://https://aclweb.org/anthology/W/W15/W15-3212.pdf

Keywords

part-of-speech tagging
natural language processing
computational linguistics
Arabic
Arabic Treebank
corpus linguistics

Access to Document

Cite this

@inproceedings{fdfcf5e40bf642048122d86938064082,

title = "Joint Arabic segmentation and part-of-speech tagging",

abstract = "Arabic has a very complex morphological system, though a very structured one. Character patterns are often indicative of word class and word segmentation. In this paper, we explore a novel approach to Arabic word segmentation and part-of-speech tagging relying on character information. The approach is lexicon-free and does not require any morphological analysis, eliminating the factor of dictionary coverage. Using character-based analysis, the developed system yielded state-of-the-art accuracy comparing favourably with other taggers that involve external re-sources.",

keywords = "part-of-speech tagging, natural language processing, computational linguistics, Arabic, Arabic Treebank, corpus linguistics",

author = "John McNaught and Shabib AlGahtani",

year = "2015",

month = jul,

day = "30",

language = "English",

isbn = "978-1-941643-58-7",

pages = "108--117",

editor = "Nizar Habash and Stephan Vogel and Kareem Darwish",

booktitle = "Proceedings of the Second Workshop on Arabic Natural Language Processing",

publisher = "Association for Computational Linguistics",

address = "United States",

note = "ACL Second Workshop on Arabic Natural Language Processing (WANLP 2015) ; Conference date: 30-07-2015 Through 30-07-2015",

url = "http://https://aclweb.org/anthology/W/W15/W15-3212.pdf",

}

McNaught, J & AlGahtani, S 2015, Joint Arabic segmentation and part-of-speech tagging. in N Habash, S Vogel & K Darwish (eds), Proceedings of the Second Workshop on Arabic Natural Language Processing. Association for Computational Linguistics, Stroudsburg, USA, pp. 108-117, ACL Second Workshop on Arabic Natural Language Processing (WANLP 2015), Beijing, China, 30/07/15. <https://aclweb.org/anthology/W/W15/W15-3212.pdf>

Joint Arabic segmentation and part-of-speech tagging. / McNaught, John; AlGahtani, Shabib.
Proceedings of the Second Workshop on Arabic Natural Language Processing. ed. / Nizar Habash; Stephan Vogel; Kareem Darwish. Stroudsburg, USA: Association for Computational Linguistics, 2015. p. 108-117.

Research output: Chapter in Book/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Joint Arabic segmentation and part-of-speech tagging

AU - McNaught, John

AU - AlGahtani, Shabib

PY - 2015/7/30

Y1 - 2015/7/30

N2 - Arabic has a very complex morphological system, though a very structured one. Character patterns are often indicative of word class and word segmentation. In this paper, we explore a novel approach to Arabic word segmentation and part-of-speech tagging relying on character information. The approach is lexicon-free and does not require any morphological analysis, eliminating the factor of dictionary coverage. Using character-based analysis, the developed system yielded state-of-the-art accuracy comparing favourably with other taggers that involve external re-sources.

AB - Arabic has a very complex morphological system, though a very structured one. Character patterns are often indicative of word class and word segmentation. In this paper, we explore a novel approach to Arabic word segmentation and part-of-speech tagging relying on character information. The approach is lexicon-free and does not require any morphological analysis, eliminating the factor of dictionary coverage. Using character-based analysis, the developed system yielded state-of-the-art accuracy comparing favourably with other taggers that involve external re-sources.

KW - part-of-speech tagging

KW - natural language processing

KW - computational linguistics

KW - Arabic

KW - Arabic Treebank

KW - corpus linguistics

M3 - Conference contribution

SN - 978-1-941643-58-7

SP - 108

EP - 117

BT - Proceedings of the Second Workshop on Arabic Natural Language Processing

A2 - Habash, Nizar

A2 - Vogel, Stephan

A2 - Darwish, Kareem

PB - Association for Computational Linguistics

CY - Stroudsburg, USA

T2 - ACL Second Workshop on Arabic Natural Language Processing (WANLP 2015)

Y2 - 30 July 2015 through 30 July 2015

ER -

Joint Arabic segmentation and part-of-speech tagging

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this