A Framework for Traceability in the Archival of Model-Based Experiments

Repeatability, reproducibility, and replicability are foundational principles of the scientific method. Models, as representations of salient aspects of a system-under-study, play a central role in these principles by enabling the documentation of assumptions, structures, and behaviours that can be systematically analysed and shared. However, contemporary experiment platforms provide limited support for capturing precise semantics, design rationale, and conceptual provenance of experiments. This paper presents a framework for recording and managing the intentional and conceptual traceability of experiments. The framework prescribes seven phases of an experimental lifecycle, from intent elicitation to reporting and archival, allowing to trace back to the intent. It integrates experimental concepts with various standard ontologies to provide precise, queryable semantics across experiment artefacts. A proof-of-concept implementation demonstrates how the framework enables structured documentation of each lifecycle phase of an experiment, knowledge-graph–based exploration, and export (to public repositories) of shareable artefacts in standard research archival formats with varying levels of traceability.