CREWS Validation Frames: Patterns for Validating Systems Requirements

This paper proposes a pattern language for socio-technical system design to inform validation of system requirements. The development of this language takes inspiration from Alexander’s pattern language for building design in architecture. It identifies different types of patterns which fulfil different roles in the requirements engineering process. This pattern-based validation approach has been operationalised in the CREWS-SAVRE software prototype. CREWS-SAVRE applies patterns to both scenarios and requirements documents to detect missing and incorrect system requirements, as well as to recommend new requirements which can improve the design of the socio-technical system. 1 Patterns in Requirements Engineering Patterns are a novel alternative technique to help us better acquire, model and validate system requirements. In simple terms, patterns enable people to reuse knowledge about old solutions to solve similar new problems. However, there is little reported research into patterns for requirements engineering, in spite of the considerable current interest in software patterns for system design and implementation (e.g. Gamma et al. 1995). Indeed, each existing requirements engineering research initiative still tends to have a singular focus on process, domain or language. In contrast, patterns for requirements engineering, as we shall see, cuts across the divisions between process, domain and language. After all, experienced engineers do not separate them when acquiring, modelling and validating system requirements. Patterns which describe the common elements of complex structures were first documented in the field of building architecture. Christopher Alexander, in his book "The Timeless Way of Building" (Alexander 1979), argues that "Beyond its elements, each building is defined by certain patterns of relationships amongst its elements." Patterns can be used to abstract away from the details of particular buildings and 1 This research has been funded by the European Commission ESPRIT 21903 ‘CREWS’ (Co-operative Requirements Engineering With Scenarios) long-term research project.