SNAP:DRGN - Standards for Networking Ancient Prosopographies: Data and Relations in Greco-roman Names

The Standards for Networking Ancient Prosopographies (hereafter SNAP) project aims to address the problem of linking together large collections of material (datasets) containing information about persons, names and person-like entities managed in heterogeneous systems and formats. In doing so we must address a number of complex issues:

How do we know whether a person in one large dataset is the same as a person with the same name in another dataset with a different format and metadata?
If person A in one database or network is asserted as being the same as person B in another network, do all the statements asserted about person A in the first database also apply to person B, and if so what status do these assertions have in the new context?
How do we record and integrate the provenance descriptions of both the data and the assertions to create an effective trust network through which we can assess the reliability of a given statement?
How do systems cope with very large amounts of data and how do we visualize the amount of information available?
How do we best manage the shift from human-assertions to computer-guided assertions as the dataset moves from human-manageable to big data systems without losing academic credibility in the processed results?
What academic statements need to be supported to allow the migration of prosopographic and onomastic data silos to interconnected and open data networks?
What lessons can we learn from working with person-like entity networks, when freed from the privacy and related ethical considerations that would affect modern social networks
These are questions of great interest, and difficulty, for Web scientists studying the contemporary world, given the proliferation of social networks, of different accounts, usernames, handles and URLs associated with the users of the Web, and of databases containing personal information about celebrities, authors, creators and historical figures. SNAP will address the more limited, and tractable, problem of datasets containing information about historical names and persons, whether these be onomastica, prosopographies, biographical collections, or merely indices of names from a digital text edition. Specifically, SNAP will take the coherent, geographically and chronologically constrained field of prosopography in the Classical world, in which there is established academic activity, as a delimited and realistically sized pilot project to explore solutions to this problem of networking person datasets. Using a selection of the most significant datasets in this sub-domain, we shall address some of the issues of integrating heterogeneous data at the identity level through access to very large but manageable numbers of entities, with the collaboration of the scholars and scientists responsible for the projects involved.

Initially working with a consortium of three collections of person/name data, SNAP will expand through additional partners and datasets from the academic, heritage, and commercial sectors in the early stages of the project. Assistance for the inclusion of new datasets will be provided by the project team, as part of the mechanism to test and refine the proposed data models, ontologies and schemas. These models build on existing work in the field (see Research Context) as well as being designed specifically for this data type, for compatibility and suitability across differently formatted collections. The model's fitness of purpose will be further tested through building distribution and visualization tools, with Web services (see WP3) on top of them; experimenting with the generation of data through computational techniques to leverage the new connections made for new research questions; and expanding our existing, combined dataset with new entities, identifications and connections drawn from inscriptions.

The combined datasets that the project will initially be focusing on contain over 400,000 identified person-like entities and a similar number of attestations, name-entities and annotations records. This number will rapidly expand with the creation of the entities and relationships required by the project model and through the addition of new material and datasets. We envision that the project will be working with over one million entities (persons, names, person-like entities) and their associated statements of relationship, reference and description. These records are not only big data in their own terms but represent a significant portion of the personal data available for the domain and period under investigation. As such this project has the capacity to transform the way that we understand and interact with the data and the related scholarship in the area.

This project will be a proof of concept work; a much larger project will be required to integrate the world of classical prosopography comprehensively, to expand further connections with historical person data from other periods and places, and to test the data models proposed against the more ambitious world of Linked Data as a whole. The project includes a significant time commitment to disseminate and publish the results, both in terms of the ontologies and schemata, guidelines for new projects wanting to participate, and historical articles on information gleaned from this research.