ICE-Theorem - End to end semantically aware eResearch infrastructure for theses


  • Peter Sefton
  • Jim Downing


ICE-TheOREM was a project which made several important contributions to the repository domain, promoting deposit by integrating the repository with authoring workflows and enhancing open access by prototyping new infrastructure to allow fine-grained embargo management within an institution without impacting on existing open access repository infrastructure. In the area of scholarly communications workflows, the project produced a complete end-to-end demonstration of eScholarship for word processor users, with tools for authoring, managing and disseminating semantically-rich thesis documents fully integrated with supporting data. This work is focused on theses, as it is well understood that early career researchers are the most likely to lead the charge in new innovations in scholarly publishing and dissemination models. The authoring tools are built on the ICE content management system, which allows authors to work within a word processing system (as most authors do) with easy-to-use toolbars to structure and format their documents. The ICE system manages both small data files and links to larger data sets. The result is research publications which are available not just as paper-ready PDF files but as fully interactive semantically aware web documents which can be disseminated via repository software such as ePrints, DSpace and Fedora as complete supported web-native and PDF publications. ICE-TheOREM combined the Object Reuse and Exchange (OAI-ORE, IONSREPORT 2008) and SWORD-APP protocols to transfer content between a content management system, a thesis management system and multiple repository software packages and looked at ways to describe aggregate objects which include both data and documents, and to represent structure withing thesis documents. This can be generalized to domains other than chemistry. ICE-TheOREM has demonstrated how focusing on the use of the web architecture (including ORE) enables repository functions to be distributed between systems for complex, data-rich compound objects.