Provenance Core

The Provenance Core describes research entities through the experimental and computational processes in which they participate. It connects resources with the activities that generate, use, or transform them.

Provenance metadata record those processes and connections. The Provenance Core provides the concepts and relationships for this information within the ARC Abstract Model, giving readers a route through the recorded history of a research output.

#Processes, inputs, and outputs

Provenance descriptions relate processes to their inputs and outputs. They can also carry methodological information, such as experimental conditions and computational parameters, that helps explain how a process was performed.

This applies across experimental and computational work. Acquiring a measurement and analyzing a measurements table are different activities, but both can be described through the resources they use and produce. Recording these connections helps keep the stages of research interpretable together.

#A shared example

Consider a dataset containing a table of experimental measurements and a derived analysis result. Provenance metadata connects the measurements to their acquisition process. It also connects the derived result to the analysis process and the measurements used as input.

A reader can follow those recorded relationships from the result back toward its origins. The detail available depends on what was documented; the existence of a connection does not, by itself, provide a complete account of the method.

#Distinguishing process context from meaning

An instrument setting can explain how a measurement was acquired. It does not automatically become a measured characteristic of the material. Similarly, an analysis parameter records a choice in processing without necessarily describing a property of the output itself.

The semantic core supplies the scientific meaning of variables, values, and units. The administrative core records identity, responsibility, and reuse conditions. Together, the three perspectives support interpretation without treating every connected annotation as interchangeable.

#Specialized descriptions and representations

The ISA extension relates research organized through investigations, studies, and assays to the common model. Workflow Run adds specialized descriptions of computational workflow execution.

Provenance can develop incrementally in an ARC Workspace, then be expressed through ARC RO-Crate for publication. Preserving the relevant process relationships is part of preparing useful context for consumers beyond the original research team.