Academia.eduAcademia.edu

Annotating Digital Collections (invited respondent)

AI-generated Abstract

The article discusses the challenges and considerations in annotating digital collections in the context of digital humanities (DH). Emphasizing the social aspects over technical ones, it highlights issues such as access to collections, funding requirements for free use, and the need for sustainability in digital resources. The author explores the complexities of integrating diverse collections, particularly when employing standards like TEI, and questions how meaningful a collection is if parts of the data are inaccessible.

Annotating Digital Collections: A Response Peter A. Stokes Dept. ASNC, Cambridge pas53@cam.ac.uk Use Cases 1. Require annotation, specifying regions, zoom 2. Require using images from different repositories 3. Both focus on images, but we also want to search across collections, etc. Image Markup, ROI, Zoom… • Need to specify regions, add annotations – Can link to zoomed region (Google) – Can annotate using TEI, etc. (IMT/TILE) – Can ‘chop out’ images (Dunhuang) • (Probably) private and public annotations – Want (need!) to use digital images/collections as evidence for academic argument – Issues of sustainability, reliability, trust… – There are ways of documenting: MIX/METS, etc. Access to Collection Content • Users may not have access to all collections • Funding bodies may require free access, deposit of data, acknowledgement... • Depend on all collections for sustainability • Content providers often require users agreeing to terms and conditions before use How meaningful is the collection if part of the data is missing? ‘Basic’ Search and Integration • XML, TEI designed to be flexible – TEI allows many ways of doing the same thing – Scholars have different conventions anyway: terminology, editorial practice, orthography, dates… – How can we combine collections which use TEI etc. in different ways? • Frameworks/standards exist (ENRICH, etc.), but don’t really work in my experience Final Thoughts We have to build collections with these uses in mind: this is what people expect in DH The problems are not too hard for one or two resources; what about five, fifteen, fifty?? The technology is (generally) easy: the problem is ‘social’