In Part 1 of this series, we provided an overview of data (or knowledge) commons and some key issues to consider, but how does one actually create and manage a data commons? To find your feet in this budding field, build on the theoretical foundation; address the specific context (including perceived objectives and constraints); deal with the thorny issues (including control and change); establish a core set of principles and rules; and, perhaps most importantly, plan for and enable change.
Foundation and Context
Remember the purposes, advantages, and limitations of a commons. Participants are trying to tackle one or more critical problems (like a spirited disease), and a data commons can be a powerful tool. Research suggests that shared values and goals and a foundation of trust increase the likelihood of success. Optimize the specific circumstances your community faces, and continue to monitor those circumstances and the corresponding best practices.
Issues and Rules
For a data commons, the standard rules and agreements will need to be agreed upon by many participants of varying levels of influence. It would be too burdensome (and costly) to negotiate customized agreements with each participant, and the other participants need some comfort that the rules are clear and standard. So consider developing some minimum standard terms that would be almost universally accepted. Though priorities and concerns differ, examples include the following:
- Data Contribution
- The contributed data must be de-identified (or even anonymized, where all links between an individual and the data are irreversibly broken). Think through any regulations (e.g., HIPAA or GDPR) that might apply.
- The contributed data must be in an acceptable format.
- The commons manager must require users to agree to the standard terms.
- The commons manager must limit access, as applicable. For example, certain types of data might be within a protected zone that is limited to certain types of users.
- The commons manager must impose additional terms that are specific to the contributed data (e.g., no downloading). This framework permits flexibility instead of imposing many rules on all data. If the data contributor changes any terms specific to the contributed data, such changes should only apply going forward.
- Data Access and Use
- Users must comply with the law, professional standards, and data commons policies.
- Users may not attempt to identify or contact any data subject.
- Users may not disclose the data to any unauthorized person.
- Users must comply with any attribution requirements associated with the data.
- Users may not sell or otherwise commercialize the data. But note that there may be strong views on both sides of this issue.
- The commons manager may use commercially reasonable tools and methods to monitor and enforce the data commons terms and policies.
- Responsibility and Liability
- The commons manager, which could be one or more individuals or nonprofit organizations, will want to limit its exposure. A manager’s liability should be limited, but some narrow exceptions might be acceptable to both the manager and the other participants (e.g., gross negligence).
Consider establishing committees with oversight responsibilities. Examples include membership, access (i.e., authentication), enforcement, and change. Committees or other groups can streamline feedback and improvement.
Measure activity and outputs. Foster communication and participation. Establish governance and change mechanisms (e.g., voting). All of the foregoing tips relate to the fact that if participants feel heard and empowered, you may notice buy-in, consensus, and transformation.
Remember that we’re still early in the process. Practitioners are working from, and integrating with, old models and forms. Scholars are still formulating theories, design principles, and guidelines. And all the while, traditional systems continue to entrench themselves in society. But the social, health, and other problems we collectively face are real and complex. We continue to be impressed by, and we are grateful for opportunities to work with, those vigorously creating (and iterating) open and effective knowledge commons.