Home Research Fanyu Wang
PhD Student · Primary Supervisor

Fanyu Wang

Requirements-Driven Software Quality Assurance using LLMs

FW

Monash University

Primary Supervisor: Dr Chetan Arora

Co-supervised with A/Prof. Aldeida Aleti · A/Prof. Chakkrit Tantithamthavorn

Research Theme: AI for Requirements Engineering →

Project illustration

Research Project

Requirements-Driven Software Quality Assurance using LLMs

Fanyu's research explores the potential of large language models to enhance the Requirements Engineering phase of software development — a critical yet challenging domain due to communication complexities and early-stage uncertainties.

The project leverages LLM proficiency to bolster efficiency and precision across all RE tasks: elicitation, analysis, specification, and validation. By placing requirements at the centre of the quality assurance process, this work aims to create tighter feedback loops between what stakeholders need and what software systems deliver.

A significant thread of this work focuses on automated software testing driven directly by natural language requirements — reducing the manual effort of translating specifications into executable test scenarios. The research is validated in industrial settings, with applications spanning the space industry and enterprise software domains.

Publications

Selected Publications.

TOSEM · A* 2025 Requirements-driven automated software testing: A systematic review
+

Wang, Fanyu, Chetan Arora, Chakkrit Tantithamthavorn, Kaicheng Huang, and Aldeida Aleti. ACM Transactions on Software Engineering and Methodology (2025). A systematic review of the state of the art in using requirements artefacts to drive automated testing — synthesising methods, tools, and open challenges in the field.

ASE · A* 2025 Multi-Modal Requirements Data-based Acceptance Criteria Generation using LLMs
+

Fanyu Wang, Chetan Arora, Yonghui Liu, Kaicheng Huang, Chakkrit Tantithamthavorn, Aldeida Aleti, Dishan Sambathkumar, and David Lo. ASE 2025. Presents a novel approach for automatically generating acceptance criteria from multi-modal requirements artefacts using LLMs, validated in real-world software engineering contexts.

IEEE RE · A 2025 From domain documents to requirements: Retrieval-augmented generation in the space industry
+

Arora, Chetan, Fanyu Wang, Chakkrit Tantithamthavorn, Aldeida Aleti, and Shaun Kenyon. IEEE RE 2025. Investigates the application of retrieval-augmented generation to extract and formalise requirements from unstructured domain documents in the space industry — an environment with stringent correctness demands.

Workshop 2025 Prompt engineering for requirements engineering: A literature review and roadmap
+

Huang, Kaicheng, Fanyu Wang, Yutan Huang, and Chetan Arora. REW 2025 (IEEE RE Workshop). Surveys the emerging intersection of prompt engineering and requirements engineering, synthesising current practice and identifying a roadmap for future research directions.

Collaboration

Interested in collaborating?

Get in Touch →