Collaborative Data Maintenance

Collaborative data maintenance

Collaborative approaches to collecting and managing datasets can be cost-effective as well as increase data quality, across public and private sector organisations, creating sustainable, trustworthy data infrastructure.

If you think of collaborative approaches to data collection, you might think of OpenStreetMap, Wikidata, or ebird. It is however quite complicated to keep these projects alive and functioning over long periods of time.

In this project, we explored and documented patterns of collaborative maintenance of data assets to help guide public and private sector organisations to successfully benefit from these approaches.

For me, the project consisted of a lot of research, taxonomy, organisation of large amounts of information, UX design, and team lead.

Information organisation: as we were a team of 5, it was important that we were all aligned on what we were trying to achieve, that we used a common vocabulary, a similar way of describing different concepts. This was done by collating information gathered from the research in workshops, to create themes and a logical organisation of each pattern (problem, context, solution, example). It also meant that we had to agree on what was important and not. Ongoing conversations and managing expectations was key in this step of the work.

UX design: the website had to be usable, navigable, helpful and convenient. Working on a landscape analysis to understand what libraries could look like, I designed a UX solution for the patterns that ticked all the boxes. Due to very limited means however, the implementation had to be simpler than I had hoped. However, the actual catalogue of patterns is convenient and understandable.

Research: it revolved around understanding why certain projects were successful while some others weren’t, what made them what they were, more or less known or popular within their sectors, how to identify the right set of data, how to talk to the collaborative teams so that they would be interested, willing to participate and keep participating.

Taxonomy: The output of the project was a patterns library explaining how to develop a successful collaborative dataset over time. It had to be put together in a way that was understandable by people new to data, usable and with several layers of actionable information. Using remote card sorts and tree tests, as well as interviews, we managed to get to an understood taxonomy.