Requirements extraction from model-based systems engineering: A systematic literature review

Abstract

Collaboration and easy data exchange are crucial in modern systems that involve hardware, electronics, software, and users. Requirement Engineering (RE) and Systems Engineering (SE) are challenging fields that require tool support to automate activities. Natural language (NL) requirement documents can create processing issues. To address these issues, detailed models have been developed to represent a system effectively. These models are intend to replace inconsistent documents over time by using model-based methodologies like Model-Based SE (MBSE). Within the MBSE methodologies, Arcadia/Capella has proven its capabilities as a comprehensive tool in the SE community to define and validate complex system architecture. Thus, this paper aims to investigate the tools, methods, techniques, or processes for extracting requirements from the MBSE environment or model generation from NL requirements. Furthermore, this discusses how these approaches are applied specifically in the Arcadia/Capella and how transforming requirements are addressed to support textual requirements. We conducted a systematic literature review (SLR) by selecting 97 articles to examine advances in this field in various aspects of these approaches. The results presented in this SLR uncovered several key findings that have important implications for future research, such as the dominance of the model generation from NL; transforming model-based requirements to NL requires more data; and the fact that requirements extraction in Arcadia/Capella needs more evidence.