Abstract
This paper aims to propose a specific formalism for Arabic morphology modeling that is too complex to model exhaustively with classical approaches. Therefore, it was necessary to find out an adequate representation of formalism. We designed, thus, a declarative, object-oriented language, referenced to us: MorphoScript, which allowed us to represent the complete morphological knowledge that we could identify optimally. The study that we are presenting here aims to propose an adequate data model of natural language morphological components and composition rules. We will thus present the basic elements and the theoretical and technical foundations of a language reproducing and assisting a morphological analysis process and the principles that guided the conception of this data model fully based on class concepts. Therefore, it is an object-oriented language using inheritance as basic support to define the morphological links between the different morphological classes. We have also used aggregation concepts and an annotation indexing system allowing the morphological designer a better representation of morphological knowledge.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
