Evolving methodology

Towards the end of Phase 1, we looked at how our methodology had changed since we conceived the project. That discussion continued during the second sandpit and indeed the methodology continues to evolve during Phase 2. To begin with, we had expected to find common practice among the project partners, but in practice we found very little, as the majority of the metadata in use was domain-specific. While true that we found Dublin Core-like annotations in all the examples, such values did not correspond to the concept of active metadata use. We therefore looked more closely at the importance of provenance, discussing in particular the roles of prospective and retrospective metadata.

During the week prior to the second sandpit, Mike Mineter attended the TaPP conference in Edinburgh and presented a poster that described not only our methodology but also our emerging understanding of what the active use of metadata was really about. Improved comprehension led us redirect our project goals to focus more on creating guidelines and exemplars to demonstrate how the active use of metadata can lead to more effective research, and much less on identifying core elements. We retained the notion of a core model but with a meaning more methodological than schematic. One intriguing idea that emerged during the BOF session at the second sandpit was that active metadata might represent, in part at least, how prospective provenance turns into retrospective provenance: the filling in the “provenance sandwich”.

In this first month of Phase 2, we have continued modelling examples with Annalist, aiming to understand how to capture – in the core model – how metadata is being used actively in each case and how that usage might be generalised to inform the guidance. We are also considering how in due course we should develop that guidance and the exemplars that will support it. At the same time we are refining our description of the concept of active use, which is a key component of our evolving methodology for producing the guidance.

Colin Bird (acknowledging the helpful notes taken by Graham Klyne during the second sandpit)