Domain-specific requirements analysis framework: ontology-driven approach
Shreya Banerjee, Anirban Sarkar · International Journal of Computers and Applications · 2019
Requirement engineering (RE) is the first track in the course of software development that represents and analyses the problem domain for suitable solutions. This improves the quality of the software development and decreases the risk of budget overrun, delays and project failure. However, traditional methods are inadequate to address different problems in RE. Prime challenges among those are lack of interoperability between heterogeneous requirements models, less focus on validation of requirements, and inability to reuse similar requirement specification in different domains. To solve these challenges, semantic approaches are needed to represent commonalities and variability in requirements across different applications and domains. This paper is aimed to address this issue. A requirement analysis framework named as Ontology-driven requirement analysis (ODRA) is proposed in this paper. Proposed ontology-driven specification provides common semantics towards different requirements. Proposed framework also comprises Formal Scenario descriptions and cause-effect-dependency graphs to conduct an early and a detailed analysis of requirements. It is capable to represent 5W1H characteristics (What, Why, Who, When, Where and How) of requirements. Further, it devises a set of key validation rules and a customized rule-based reasoner to validate consistency, unambiguity, and traceability in requirement specifications.