Authors
Steve Bennett, Judith Pearce, and Nick Nicholas (Link Affiliates)
Abstract
We demonstrate a method of analysing infrastructure requirements in e-Research in terms of generic functionality, challenging the perception that every discipline and institution’s requirements are unique. The need to capture common infrastructure requirements across disciplines and meet them with reusable services is recognised by both Project Bamboo and the e-Framework; we combine the two approaches with a novel modelling technique to facilitate the combining of discrete blocks of functionality.
Project Bamboo, a wide-ranging American e-Scholarship initiative to improve the use of digital technologies in humanities research, has focused on this goal of identifying common requirements – expressed as “scholarly narratives” and “recipes” – and matching them with common services. We advance towards this goal by employing the e-Framework, a methodology for orchestrating generic technology services to support related business processes, in “service usage models” (SUMs). We develop the e-Framework’s notion of combining SUMs, using UML notation to define how SUMs are connected. This allows SUMs representing discrete blocks of generic functionality, such as content syndication or annotation, to be assembled into larger systems, such as a collaborative biographical encyclopaedia.
Specified this way, these systems are independent of discipline and technology. This specific example is motivated by the Australian Women’s Archive Project, in the University of Melbourne’s e-Scholarship Research Centre. With AWAP’s assistance, we analysed their principal repository, the Australian Women’s Register, contributing “scholarly narratives” and “recipes” to Project Bamboo. We thus propose the discrete, generic groupings of services that collectively support a collaborative biographical encyclopaedia based on a searchable collection with syndication, reviewing, annotation and online contribution.
This work has important implications for building sustainable, shareable, evolving collections, and more generally efficiently deploying software infrastructure and increasing service reuse in e-Research.
About the speaker
Steve Bennett is a business analyst with Link Affiliates, a team of consultants working in standards and interoperability projects in education and research. In this role he has been using the e-Framework to facilitate services analysis. With degrees in software engineering, French and linguistics, his background includes technical writing, data migration and software build engineering. He has previously worked on the ARCHER project, for a database integration vendor in France and a secondary school management system vendor in Melbourne. His other interests include usability, text parsing, and free content, notably Wikipedia.