Supporting and enhancing your repository: A joint RepNet/RSP workshop

A joint RepNet/RSP workshop on enhancing Institutional Repositories (IRs) and their metadata sets was held at the BCS in London on January 21st, 2013.  The event provided an overview of the UK RepositoryNet+ Project current status and aims for the future, plus presentations on the just-released draft RIOXX Metadata Application Profile and on the CrossRef set of APIs for enabling automatic metadata collection into IRs. CrossRef ‘DOI Lookup’, and FundRef features will be looked as part of the developing RIOXX application profile guidance.

A series of breakout sessions were held after the introductory Jisc Repository and Curation Shared Infrastructure (RCSI) and RepNet presentations by Balviar Notay, Andrew Dorward and Pablo de Castro. These were aimed at getting the event attendees’ – a mixed community of IR managers, research information managers, project managers and developers, see event attendee list – input into what functionality they would like to see as part of the RepNet Applications/Use Cases.  This was a chance for the community to feed in their requirements on how national services could improve local workflows and to discuss the added value the RepNet could provide for Repository Managers.

Four breakout sessions were held on (i) reporting and benchmarking, (ii) deposit, (iii) metadata enhancement and (iv) repository platforms. Each of the groups had a pre-established set of issues to deal with (see figure below), plus clear instructions from the Jisc RCSI Programme manager to focus on infrastructure-related aspects.

 

Image

 

IRUS-UK Project was the focus of the Reporting and benchmarking discussion group, with examination of its present features and its potential for further reporting from RIOXX extended metadata set and through a future aggregation service. Enhancements to IRUS were suggested such as coverage extension to other types of materials beyond articles (e.g. learning resources) and additional IR metrics (among them metadata-only to full-text item ratio, Gold vs Green OA items or top monthly downloads by IR and nationally) that would provide an improved cross-repository description both for benchmarking purposes and as a useful feed to SCONUL. Finally, (P)IRUS prospects of using cross-repository COUNTER-compliant stats to benchmark with at least one publisher site were agreed to be important for demonstrating the value of the IR network for scholarly communications.

Discussion on Services Supporting Deposit addressed technical and content requirements for the RJ Broker, such as how better metadata could improve the deposit process between scholary systems; SHERPA RoMEO service possible enhancements (together with JULIET); usefulness of Open Access mandates for content collection and mechanisms to make them more effective, and the need for publisher involvement in defining clearer deposit policies.

The call for RoMEO enhancement issued at this session was to a great extent shared by all discussion groups and included the need for harmonisation in publisher policies for deposit – with suggestions for standard, simplified agreements to be put in place – plus coverage of copyright licenses (CC BY vs CC BY NC), journal compliance with specific Open Access policies and eventually article processing charges (APCs) rates at journal level. A wish was also expressed for clarification of Gold Open Access publishers’ rights regarding deposit into IRs.

The breakout session on Metadata Enhancement looked at aspects of metadata quality and their impact on the provision of repository services. While the breakout group was characterised by a focus on repository management – and viewed metadata quality primarily in terms of its ability to support this – the discussion did range into aspects of ‘discovery’ through the exposure of rich metadata. Specific aspects of metadata quality enhancements discussed included:

  • a desire to be able to consistently describe resources according to their publication status according to a standard vocabulary (e.g. pre-print, accepted for publication, final copy etc.).
  • a mechanism for indicating versions of items
  • the perennial problem of precisely identifying authors
  • approaches to linking metadata records to records of related resources (e.g. from a scholarly paper to a research data-set which it references)
  • getting better quality metadata about other literature types (e.g. books, conference proceedings).

Other topics discussed included the RIOXX guidelines and metadata application profile, the FundRef project pilot, and ORCID. The group was interested in each of these and was able to offer feedback on these various initiatives. A quick ‘straw poll’ of the group suggested that intent to engage with ORCID was widespread among the institutions represented.

Breakout session 4 on Repository platforms put together a list of already available IR enhancements that should ideally become more widely implemented, including SWORD endpoint availability, IRUS-UK, OpenAIRE compliance, KeepIt! preservation feature or CRIS and REF2014 plugins for EPrints among others. The issue was raised that IR managers are often unaware of the different services available – something the RepNet Survey has recently provided useful evidence about. Since technical challenges for feature implementation are not critical enough to explain low implementation, this was mainly judged to be a result of the severe lack of awareness-raising mechanisms regarding available services. A request was made as a result for (i) investigating the appropriate hierarchical level at HEIs for proposing new IR service implementation and (ii) providing RepNet factsheets, ie one-page summaries providing human-understandable info about technological challenges for a given service plus best practice in its successful implementation by a specific IR manager.

IRUS-UK service, whose patch takes 10 minutes to install for EPrints-based IRs and 30 minutes for DSpace-based ones, was identified as a potentially useful testbed for outlining successful strategies for widespread enhancement adoption by IRs – including the possibility of having its patch for EPrints added to the EPrints Bazaar service catalogue. Up-to-date adoption snapshots such as the RepNet-featured IRUS implementation map were also agreed to be useful.

A presentation by Paul Walk followed the breakout sessions focusing on the first draft release of RIOXX application profile v0.8, which is now open to consultation. An overview of CrossRef features for automatic metadata collection delivered by Geoffrey Bilder provided a useful follow-up to this discussion. The afternoon session also featured Yvonne Budden on IRs taking centre stage in the future for the OA publishing activities at HEIs.

This event highlighted that there was much dependency on having  accurate and consistent metadata in repositories to then allow  robust added value services to be built on top of them. The RIOXX work and initiatives like CrossRef and Fund Ref are starting to address the gaps in this area. The breakout sessions provided evidence as well for the fact that the repository community may now be reaching some consensus on what the relevant topics to address should be. Obviously more work will be required to validate these national services with further community engagement.

All presentations are available at the RSP website, and a storify story includes the workshop tweets. 

 

One Response to Supporting and enhancing your repository: A joint RepNet/RSP workshop

  1. Re: “Enhancements to IRUS were suggested such as coverage extension to other types of materials beyond articles (e.g. learning resources”.

    IRUS-UK already covers all item types in the participating IRs. They are mapped to a limited number types (The [v1.0] IRUS list contains the following item types: Article; Audio; Book; Book Section; Conference or Workshop Item; Exam Paper; Exhibition; Image; Journal; Learning Object; Moving Image; Other; Patent; Preprint, Report; Review; Text; Thesis or Dissertation; Unclassified; Website; Working Paper. A [v2.0] list has been defined, but not implemented yet.)
    So they are counted, but cannot be reported on ‘COUNTER-style’ following the Code of Practice recently announced (http://www.projectcounter.org/pirus.html)

    Also, the suggested “additional IR metrics” of “top monthly downloads by IR and nationally” is and always was already provided.

    Re ‘DOI Look-up’. CrossRef were a partner in the PIRUS2 project and the following
    is taken from the PIRUS2 Final Report:

    “It is possible, in many cases, to use those to query and retrieve the DOI from:
    • The CrossRef database using a variety of synchronous and asynchronous methods
    • The PIRUS2 database – where an article has been previously identified and is already known to the system

    The query matching rate is dependent on a number of factors, including comprehensiveness of metadata in the CrossRef and PIRUS2 databases, matching technology, and data quality. It is difficult to put an exact figure on the success rate of author/title matching (the more sophisticated the matching algorithms, the higher the success rate) but we estimate that, realistically, matching can work in 85%+ of cases.”

    Thanks for the other suggestions!
    However, what would be really good would be if all IRs participated with the existing service ASAP!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: