Evaluation of dependency parsers for long Arabic sentences

Maytham Alabbas, Allan Ramsay

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The aim of the work reported here is to investigate the effectiveness of dependency parsing for the analysis of Arabic. Arabic has a number of characteristics, described below, which make parsing it particularly challenging. The results of our investigations suggest that dependency parsing can produce reasonably accurate results. We show in particular that combining the output of two different parsers can produce more accurate results than either parser produces by itself. © 2011 IEEE.
Original languageEnglish
Title of host publication2011 International Conference on Semantic Technology and Information Retrieval, STAIR 2011|Int. Conf. Semant. Technol. Inf. Retr., STAIR
PublisherIEEE
Pages243-248
Number of pages5
ISBN (Print)9781612843537
Publication statusPublished - 2011
Event2011 International Conference on Semantic Technology and Information Retrieval, STAIR 2011 - Putrajaya
Duration: 1 Jul 2011 → …

Conference

Conference2011 International Conference on Semantic Technology and Information Retrieval, STAIR 2011
CityPutrajaya
Period1/07/11 → …

Keywords

  • Arabic Treebank
  • Dependency parsing
  • Dependency structure
  • Phrase structure

Fingerprint

Dive into the research topics of 'Evaluation of dependency parsers for long Arabic sentences'. Together they form a unique fingerprint.

Cite this