Automatic Treatment of Arabic Sentence Based on Linguistic Levels Integration: the Case of Complementizer

Abstract

The aim of the paper at hand is to succincly account for multiple syntactic
and semantic issues at play in the automatic treatment of Arabic syntax.
Drawing on various integrated linguistic levels, major focus is placed
on the structure of the interpreted complementizer that occupies either
subject or object positions. For example, «تنجــح أن أحــب «can be interpreted
.«يعجبنــي نجاحــك» as interpreted be can» يعجبنــي أن تنجــح» while ,»أحــب نجاحــك »as
The paper approaches the issue of complementizer from the perspective of
syntactic and morphological agreement, on the one hand, and syntactic and
semantic comptability, on the other. The objective of the study is to explain
these levels’ roles in accounting for the linguistic aspects that present real
challenges in automatic processing of Arabic syntax. The challenges are
apparent throughout the various phases of analysis on which the analyzer
we suggest is built, starting from the morphological level, then the syntactic
level, to the lexical level. In this context, a point to be made is that the
morphological level provides the analyzer with the morphemes required
for linguistic units construction; the syntactic component is supported by
a set of syntactic rules and patterns to facilitate the analyzer performance,
whereas the combinatorial lexicon provides the syntactic and semantic
features.