SEEK: Salford Environment for Expertise and Knowledge

Journal Article (Refereed)
October 2008

DEAR-ONTO: A DErivational ARabic Ontology Based on Verbs

Belkredim, F & Meziane, F 2008, 'DEAR-ONTO: A DErivational ARabic Ontology Based on Verbs', International Journal of Computer Processing of Languages, 31(3), pp.1-13.

Abstract

 The Arabic language is a highly derivational and inflexional language in which

morphology plays a significant role. However, a large number of the derivations
obey a set of pre-defined rules and when applied, they do not only produce a set
of related words but also associate contextual and semantic knowledge to these
words. In this paper, we describe the Meta-Ontology DEAR-ONTO, A DErivational
ARabic Ontology used to model the Arabic Language. The Meta-Ontology uses
verbs as roots and identifies the various classes formed by the derived words.
Some of the possible applications of the DEAR-Ontology are discussed.

Notes

 There was a lot of interest on Arabic Natural Language Processing over the last two decades. Most of the research concentrated on the development of lexical resources, morphological tools and named entity recognition systems. To reach the same level of maturity as other language such as English and French, more robust tools based also on semantics should be developed for the Arabic language. The work presented in this paper is the first attempt to define words and categories in Arabic using trilateral verb roots which is the basis on which the language is developed.

 

Authors

SEEK Members

External Authors

Fatma-Zohra Belkredim

Publication Details

Journal Name
International Journal of Computer Processing of Languages

Volume
31(3)

Pagination
1-13.