HAZOP-based approach to pattern identification for non-functional requirements

Frequently requirements are written in natural language which may cause communication problems-misunderstandings. Misunderstandings are sources of risk that can severely impact the whole software development project. Since patterns and templates are tools for improving the quality of requirements, the question arises whether they can be treated as risk guards. In the paper we propose a method based on HAZOP for pattern identification for non-functional requirements (NFRs). Our exploratory case study on cost and effectiveness of the proposed method included 125 NFRs. The analysis of the identified problems led us to propose three process patterns and provided rationale for using a three templates. Generally, the proposed method might be used to improve the existing patterns and templates as well as to develop new ones.