Create a multi-tiered framework for certification of repositories
From GRDI2020
This is a GRDI recommendation; return to Main Page with all the challenges or to recommendations
Context and challenges
The challenge addressed by this recommendation is related to Data Security, Data Provenance and Trust as well as Funding, Sustainability and Governance. Audit and certification of data management is the basis upon which trust between actors can be built, including users, data centers and commercial service providers. Besides the potential in ensuring quality of repositories and data, it may also trigger a healthy convergence of systems and practices.
However, various things can be certified and a multi-tiered certification framework is needed. This will ensure that the following issues are addressed individually as well as their interaction:
- Bit Preservation (cf. Data Curation and Preservation)
- Research data (cf. the partner recommendation implement and promote best practices for data curation as part of good scientific practice)
- Data repository (i.e. the overall OAIS system)
There are various initiatives focusing on audit and certification of research data repositories. These include
Recommendation
Create a multi-tiered framework for certification of repositories that (a) interacts with requirements for data management from diverse communities, (b) builds on the experiences from existing data repositories to ensure feasibility and sustainability, and (c) feeds into existing standards-building initiatives.
One should create a taxonomy of use cases that define the different levels of requirements in terms of the importance (e.g. direct impact on policy) of the data repository. For example, for the highest ("IPCC-level") level one would most likely require open access to data and automatic reproduction of results.
One should also leverage community to develop the framework, data impact assessment and monitoring purposes. I.e. collect comments, monitor level of use & usability, facilitate fraud detection.
Stakeholders and Impact
More or less all individuals and entities involved in data-related activities are stakeholders. The impact of better understanding of different levels of certification requirements in different use cases will allow better targeting of effort and funds for high-level certification where it is needed while allowing the less refined data still to be stored cost effectively.
The certification framework will also indirectly support evidence-based policy discussion since the validity of the argumentation and the underlying data can be discussed using a common reference model.
