Title Prompt engineering for facilitating requirement elicitation from multi-format inputs
Authors Razinskas, Mantas
Full Text Download
Is Part of CEUR workshop proceedings: joint proceedings of the BIR 2025 Workshops and Doctoral Consortium co-located with 24th international conference on perspectives in business informatics research (BIR 2025): 17-19 September 2025, Riga, Latvia / edited by M. Henkel, R. Pirta, M. Kirikova, P. Forbrig, ... et al... Aachen : CEUR-WS. 2025, vol. 4034, p. 236-245.. ISSN 1613-0073
Keywords [eng] requirements engineering ; large language models ; model-based development ; requirements elicitation
Abstract [eng] Requirements Engineering (RE) increasingly relies on semi-structured and multi-format inputs such as textual descriptions, stakeholder notes, wireframes, and technical specifications. This paper proposes a hybrid methodology that combines structured prompt engineering techniques, a conceptual metamodel, and a conceptual agent-based architecture to facilitate requirements generation from heterogeneous sources. The approach operationalizes independent, negotiable, valuable, estimable, small, and testable (INVEST) framework through prompt templates and applies a metamodel to formalize the transformation process, enabling evaluation and refinement of outputs. The conceptual architecture introduces modular agents for extraction, validation, and refinement tasks, facilitating collaboration and continuous improvement. Initial experiments demonstrate that prompt-based elicitation could help in eliciting requirements. The methodology addresses gaps in existing Artificial Intelligence (AI) and model-based RE approaches and contributes a framework for integrating Large Language Models (LLMs) into Agile requirements workflows. Future work includes prototyping, large-scale validation, and domain-specific adaptation.
Published Aachen : CEUR-WS
Type Conference paper
Language English
Publication date 2025
CC license CC license description