Natural Language Processing and Formal Concept Analysis Technologies for Automatic Building of Domain Model

M.G. Ilieva and O. Ormandjieva (Canada)


Knowledge Representation, NLP, AI approach to Requirements Engineering, FCA, Domain Model


The challenge in the creation of RE/SE (Requirements Engineering/Software Engineering) models automatically from NL (Natural Language) description of requirements is to discover the knowledge model within the language model. In this article, we present a new approach for analyzing and processing these two models. Our approach combines two technologies: The first is NLP (Natural Language Processing) which we use to construct a graphical model of the language and of the knowledge it simultaneously carries within it. Once the model has been built, we can extract structural analogies with a DM (Domain Model). The second is FCA (Formal Concept Analysis), with which we process these structures. FCA helps us in two ways: as an analytical tool to formalize the concepts, and as a technology to structure and visualize them. A case study is provided to demonstrate the applicability of our approach.

Important Links:

Go Back