Natural Arabic Processing for Language Understanding Systems


Project info


NAPLUS is a prototype system that aims at two basic objectives:
  1. the processing of natural Arabic language and
  2. producing a linguistic analysis ('understanding') of the processed text.
The processing of the input language includes morphological, grammatical and semantic aspects. The outcome is a representation of the Arabic language that enables users to have multilingual interface. Future uses of the NAPLUS system is envisaged in automatic translation, database front-end query systems, Internet interface, languages teaching and self learning packages, multilingual organisations such as the United Nations, its related organisations, and the European Community.

The Arabic language is a derivation based language in which etymology and morphology play a significant role. A morphological analyser forms the core of the system. The approach to building such an analyser is based on a mixture of two approaches, a mathematical combinatorial one and a linguistic rules definition.

A basic component of the system is the Lexicon, which is accessed by all modules and interfaces to the system in order to produce the semantic representation of the parsed text. The lexicon is based on the semantic units and classifies the language according to meanings rather than according to roots. The lexicon is based on concepts and subjects related to the Semantic Field notion, or what is called Lexical Field, implying that vocabularies will be grouped under one notion or concept. Conceptual dependency and case grammars are among the basic theories that will be investigated and used accordingly keeping a very close eye on the speciality of the Arabic language and the suitability of such approaches to be applied on the language.

Updated: 13-02-2008 Contact NAPLUS