Supporting Change Impact Assessment with LLMs
Assurance Cases (ACs) are structured arguments intended to show that a system is safe. ACs require continuous maintenance through their life cycle, and manual review is often required to determine how an AC may be impacted by different update scenarios. This is expensive and error-prone as ACs can be large and complex. In this vision paper, we present challenges and opportunities for leveraging large language models to support change impact assessment for ACs.
Authors
- Claudio Menghi (ORCID: https://orcid.org/0000-0001-5303-8481)
- Logan Murphy (ORCID: https://orcid.org/0009-0001-1150-5704)
- Simon Diemert (ORCID: https://orcid.org/0000-0001-9493-7969)
- Torin Viger (ORCID: https://orcid.org/0009-0000-2611-8670)
- Marsha Chećhik (ORCID: https://orcid.org/0000-0002-6301-3517)
Institutions
- University of Bergamo (IT)
- University of Toronto (CA)
- Critical Systems Labs (CA)
Publication Details
- Published
- 2024-10-28
- DOI
- https://doi.org/10.1109/issrew63542.2024.00079
- Citations
- 2
- Primary Topic
- Big Data and Business Intelligence
- Type
- article
- Field-Weighted Citation Impact
- 1.06