Automated Generation of EXPRESS-G Models Using NLP

Abstract

The EXPRESS notation helps in representing a model’s elements, definitions, connections and cardinalities textually and graphically. The graphical version of EXPRESS language is called EXPRESS-G. EXPRESS-G provides a standardized way of representing the structure of an information model in a more reasonable way. In the recent times, the graphical representations of the software architecture are preferred due to its ability of easy to understand. In this article, a novel idea is presented to generate EXPPRES-G models from Natural Language (NL) text such as SBVR specification. In this study, an automated system is introduced that generates an EXPRESS graphical data models from SBVR, a controlled natural language. The approach is tested on various case studies and the EXPRESS models are generated from SBVR rules. Since, model representation is helpful to develop large scale systems like, Aerospace, medical science and other industries where the representation of system working is important, there is vast application of the idea discussed in this paper. This work is helpful to the data mangers and IT managers to represent their organization structure in a simple way.