2019/08/13 by Andreas Vogelsang, Markus Borg, Vogelsang, Andreas +1 · 2 citations
Business, Management and Accounting · Computer Science · #Big Data and Business Intelligence #FOS: Computer and information sciences #Machine Learning (cs.LG) #Software Engineering (cs.SE) #Software Engineering Research #Software Engineering Techniques and Practices
paper · pdf · doi:10.48550/arxiv.1908.04674
openalex publication_date 2019/08/13 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Machine learning (ML) is used increasingly in real-world applications. In\nthis paper, we describe our ongoing endeavor to define characteristics and\nchallenges unique to Requirements Engineering (RE) for ML-based systems. As a\nfirst step, we interviewed four data scientists to understand how ML experts\napproach elicitation, specification, and assurance of requirements and\nexpectations. The results show that changes in the development paradigm, i.e.,\nfrom coding to training, also demands changes in RE. We conclude that\ndevelopment of ML systems demands requirements engineers to: (1) understand ML\nperformance measures to state good functional requirements, (2) be aware of new\nquality requirements such as explainability, freedom from discrimination, or\nspecific legal requirements, and (3) integrate ML specifics in the RE process.\nOur study provides a first contribution towards an RE methodology for ML\nsystems.\n