Linked data provides methods for publishing and connecting structured data on the web using standard protocols and formats, namely HTTP, URIs, and RDF. Much like the web of documents, linked data resources continuously evolve over time, but for the most part only their most recent state is accessible. In order to investigate the evolution of linked datasets and how changes propagate through the web of data it is necessary to make prior versions of such resources available. The lack of a common ``self-service'' versioning platform in the linked data community makes it more difficult for dataset maintainers to preserve past states of their data themselves. By implementing such a platform which also provides a consistent interface to historic dataset information, dataset maintainers can more easily start versioning their datasets while application developers and researchers instantly have the possibility of working with the additional temporal data without laboriously collecting it on their own. In this paper, we describe a basic model view for linked datasets and a platform for preserving the history of arbitrary linked datasets over time, providing access to prior states of contained resources via mementoes.
