Constructing a knowledge-based quality attributes relationship matrix to identify conflicts in non-functional requirements

Abstract

A critical success factor in requirements engineering involves recognizing conflicts in Non-functional requirements (NFRs). The recent approaches use Quality attributes relationship matrix (QARM) to identify conflicts in NFRs that represents how one quality attribute undermine (-) or support (+) achieving other types of quality attributes. However, the static QARMs are not always obvious and may vary in the context of available standards, development and the involved stakeholders. In addition, these matrixes do not contain relations for emergent quality attributes viz. Recoverability, context awareness, mobility among others. Furthermore, identifying the conflicts in NFRs using the matrix without knowing the purpose of NFRs in the system may produce the false conflict identification. Hence, the aim of our research is to identify relations and influences between NFR attributes (a.k.a. Quality attributes) from available unconstrained natural language documents through automated natural language processing and machine learning that helps to deal with false conflict identification. © 2020 American Scientific Publishers.