FAIR Synthesis: Design and Technical Issues

This webpage has been archived. Its content will not be updated. View web retention policy

Many of the FAIR projects have explored particular design and technical issues in relation to their work with repositories.  Specific work undertaken by projects in the three key areas of designing the repository, harvesting and searching are described in this section.  Developments linked to these or addressing different aspects of the FAIR projects are described in related themed sections - please refer to these for further information.

Designing e-Print repositories

Several FAIR projects have documented the process they went through to design and set up their e-print repositories.  The papers below illustrate the types of decisions they made:

SHERPA also has several pages guidance on setting up e-print repositories

Metadata harvesting

Institutions that develop repositories disclose the resources to their own institutional users.  A key objective of the FAIR programme is to explore how the resources held in repositories can be disclosed to the wider community, and the OAI-PMH standard enables this.  ‘Service providers’ harvest metadata from individual OAI-compliant repositories, so users can search across many repositories.  The following documents provide an introduction to the area of metadata harvesting.

ePrints UK has developed a pilot service harvesting metadata from subject-based and institutional repositories and make them available via the RDN. By May 2005, over 65,000 metadata records had been harvested from 30 OIA-compliant repositories. The technical report below describes the design and architecture of the service and how it worked out in practice. The two issues papers highlight metadata harvesting issues and pose possible solutions.

HaIRST is harvesting metadata from a consortium of FE/HE institutions. An interesting aspect of the design is that three of the test repositories have been set up using the OAI Static Repositories specification. These static repositories are XML files that can be disclosed for harvesting by an OAI service provider. The use of static repositories may be a possible solution to allow the disclosure of XML-based content via OAI at institutions that are unable to maintain a full local repository.

Harvesting the Fitzwilliam has investigated metadata harvesting in the context of museum objects.  This experience is now being used by the Fitzwilliam Museum to facilitate data provision to a number of other OAI harvesters through the 24Hour Museum pilot and the BRICKS project.  The AHDS and ADS have also benefited from the experience of testing the harvesting of such data and are building this into how they intend to provide service access to such collections.  Harvesting the Fitzwilliam collaborated with other projects in the Museums and Images cluster of FAIR projects to develop an issues paper on images and harvesting.

Searching

DAEDALUS is using the PKP OAI Harvester from the Public Knowledge Project at the University of British Columbia to provide access across the repositories above. The PKP OAI Harvester allows you to create a searchable index of the metadata from OAI-compliant archives and features the ability to export search results into bibliographic reference management software. DAEDALUS is also planning to implement Google Scholar, and will report on their experience in a future output. 

ePrints UK and HaIRST have both used ARC harvester software from Old Dominion University. ePrints UK has developed a simple and an advanced search feature for use with their pilot service.  The simple search interface lets the user search the ePrints UK database matching terms against all available fields.  The search interface is based on WebCheshire which has been modified to include some of the fields specifically relevant to e-prints, e.g. the authors/contributors list.  The advanced search feature allows users to limit their search to specific repositories.  This feature was added in response to requests from institutions that have multiple repositories, e.g. within different departments.  This feature allows them to identify records that appear in more than one repository.

Projects in the Museums and Images cluster group collaborated to develop an issues paper on cross-domain searching.  This paper examines the issues surrounding the provision of access to multiple domain collections using OAI.

Bookmark and Share