Construction of Ontology for Software Requirements Elicitation

Abstract

Background/Objective: Elicitation of requirements from informal descriptions remains a major challenge to be accomplished in software industry. Methods: An important task in order to accomplish this goal is to construct an ontology consisting of set of concepts i.e. entities, attributes and relations based on the application domain of interest. The ontology constructed here represents the domain knowledge and requirements are the specialized subset of it. As standard description formalism the ontology is encoded using OWL DL, supported by Pellet reasoned to check the consistency of the components of the ontology. The populated ontology can be queried for matching words using SPARQL. Findings: In software development projects, voluminous unstructured text documents from different stakeholders are to be analysed and to be converted into structured requirements. This process of elicitation will be time consuming if it is to be performed manually. Domain specific ontology helps in automating the process of requirements elicitation, this article intends to construct such domain specific background ontology. The findings are elaborated in Section 3. Improvements: This article portrays the construction and use of domain specific background ontology containing the concepts and their relationships in the Automated Teller Machine (ATM) operations domain to guide the process of automation of elicitation of requirements from informal descriptions or unstructured text, which otherwise would be time consuming if carried out manually.