Exploiting ontology to map requirements derived from informal descriptions

Abstract

Requirements are narration of the services which a software system should make available along with the constraints that should be satisfied when the system operates. Software requirements have to be arrived from descriptions that are often incomplete, inconsistent, informal and ambiguous. Such informal descriptions have to be pre-processed and information constructs have to be extracted. This article deals with use of an ontology specific to automatic teller machine (ATM) operations domain that contains the concepts, the relationships that exists among the concepts and the focus is to decide on the feasibility of the requirement by mapping the extracted requirement with the requirement defined in the background ontology. The developed ontology is queried using simple protocol and RDF query language (SPARQL), if the derived requirement is present in the ontology it is said to be feasible; else decision may be taken to eliminate the requirements that are invalid and infeasible. Ontology is a formal specification of concepts with their attributes and relationship in a selected domain. As standard description formalism, the web ontology language (OWL) derived from resource description framework (RDF) is used.