Data persistence and access architecture for open and reproducible science

Reproducible science benefits the research community by providing users with credible data, which promotes accelerated progress in methodological research and innovation. This work proposes a preliminary method using persistent identifiers (PIDs) to reproduce research results using time-series data. PID is a long-lasting reference that allows specific data to be queried and accessed from a request originating from outside the system. A PID prevents unintentional alteration or corruption. The methodology and design for the prototype are proposed and built based on the IEEE Standards Association P2957 Data Governance and Metadata Management Working Group (BDGMMWG). A prototype is presented that implements parts of the Federated Metadata Registry diagram from BDGMMWG. Datasets are used to demonstrate (i.e., the prototype) how the proposed PID system architecture can be used for time-series data to create persistent links to specific data within a dataset. Providing such links are key to retrieving the data used to reproduce scientific experiments and verifying published results.