SNAP is part of an ecology of interlocking Linked Open Data initiatives, sometimes referred to as the Graph of Ancient World Data (GAWD). A forthcoming special issue of ISAW papers gives a helpful overview of the breadth of this community and its interests. One of the strengths of this approach is that each project contributes to and benefits from activities elsewhere and thus the whole becomes greater than the sum of its parts. One of these projects, Pelagios, plays a complementary role to SNAP by connecting online documents to places, rather than people. Pelagios has now been running since February 2011 – this blog post provides a little background to it and perhaps even indicate possible future directions of travel for SNAP.
The initial concept of Pelagios was born of multiple previous research initiatives, including Hestia, Google Ancient Places, EuropeanaConnect and PhD research into archaeology and the Semantic Web. It was clear that linking place references in documents to a central gazetteer was both feasible and useful at a local level, but could it be carried out at Web scale in an open and decentralised fashion? In order to achieve this Pelagios sought to produce a data framework that re-used and recycled semantic web components and online data that were already available, rather than reinvent them. Principal among these were the Pleiades online gazetteer which provides URIs for places in the ancient world, and Open Annotation, which offers a simple but flexible RDF ontology for associating information with fragments of online content.
Openness and decentralisation were achieved by helping partner projects from archaeology, classics and the wide heritage sector to link place references in their own content to the Pleaides gazetteer, saving this information in a way that could then be harvested by third parties for analysis or additional functionality. A demonstrator API was also created and a number of projects have made direct us of it, allowing them to contextualise their own data with information from that of others (and vice versa). Currently there about 830,000 annotations created by the Pelagios project and the number continues to grow.
Originally funded by JISC, the latest phase of the Pelagios has been financially supported by the Andrew W. Mellon Foundation in order to produce annotations of Early Geospatial Documents. These are both textual and visual representations of the world and its regions up to 1492, including those from ancient, Christian , Islamic and Chinese traditions. Our intention in doing so is to expand the framework sufficiently to encourage new partners to get involved from a much wider temporal and geographic scope.
What lessons might any of this hold for SNAP? In some ways SNAP faces a greater challenge than Pelagios because less of the Linked Data infrastructure is in place at the point of departure. In particular, no comparable service currently exists for ancient people comparable to that provided by Pleiades for places (although VIAF, offers a possible template for such services). This means that SNAP will need to facilitate the creation of these components, as well as the annotations which connect third party resources to them, while simultaneously seeking to avoid it collapsing into a heavily centralised service that loses the benefits of cross-community collaboration. One of the goals of this week’s workshop will specifically be to keep the various tasks SNAP seeks to support distinct so that they can each play a clear role in the GAWD ecology.