Proposal for publisher engagement with IRUS-UK

This use case outlines potential user requirements and technical development considerations for publisher engagement with the monitoring of usage statistics for Open Access (OA) resources. Building on the RepNet Wave 1 development of Institutional Repository Usage Statistics (IRUS-UK), this use case focuses on increasing engagement of publishers with testing and reviewing of aggregated usage statistics for resources made available both via publisher networks and in institutional repositories.


PIRUS and PIRUS2 were two JISC-funded projects (running, respectively, from August to December 2008 and October 2009 to December 2010) for handling Publisher and Institutional Repository Usage Statistics (PIRUS). The subsequent IRUS-UK service is a RepNet Wave 1 component for repository usage analysis. The IRUS service will provide cross-repository usage data for the whole repository network in the UK, aggregating usage statistics for items of which there are multiple copies deposited into different repositories in the network.

This use case aims to engage OA-friendly publishers for testing usage statistics for the publisher vs the repository network for a wide set of items using IRUS-UK for the purpose. Besides testing IRUS-UK integration features, this will enable a preliminary comparison of publisher vs IR usage figures for a given publisher and help make the case for IRs within institutions and also for publishers. This would ideally result in a better cooperation between publishers and institutions. (Note: see the European PEER Project conclusion on how IRs do not harm publishers but on the contrary increase their content usage.)

Example of aggregated usage statistics


  1. IRUS-UK (RepNet Wave 1 component)
  2. RJ Broker (RepNet Wave 1 component)
  3. Cooperating publishers (some of them have already agreed to take part in preliminary tests)


The feasibility of this use case depends very much on the progress of the IRUS-UK service. From the current rate of progress and given it is only aimed to be tested with one or two publishers, it seems perfectly doable within the RepNet time schedule. It may be wise, however, to give this use case a low profile.