A Distributed Canvas Rendered from Linked Data Annotations

The SharedCanvas data model enables the construction of views by distributed collaborators, by annotating a shared "Canvas" resource which is then rendered using a presentation system. This allows any sort of resource to be included from different servers or repositories, for further annotations such as commentary to be added using the same underlying technologies, and to promote further additions by third parties.

The initial use case for SharedCanvas comes from the medieval manuscript domain, where the view is of a single page constructed potentially from several images from different libraries or museums, as manuscripts are tragically often cut up and physically separated. Many different images of a page may exist, along with a transcriptions, editions and translations of the text. In other cases, the text may be well known but no images are available because the original page has been lost.

Other uses include building up canvases that depict historical newspapers, scientific articles with embedded datasets or visualizations, letters, maps or any multi-media, multi-component display. It is particularly suited to depictions of historical text, where the original physical item is of cultural importance.


2012-02-17: Panning and Zooming demo created at Stanford Code Sprint
2011-11-22: Ghent, SDH11, Books in Browsers slides available
2011-10-19: SHD11 Paper, plus September presentation slides available
2011-08-01: Initial Demonstrations available!
2011-06-10: Initial web site, separated from DMS Tech.