Repositories for research and teaching/learning material: The debate continues at #rpmeet


Last week I attended the JISC Repository and Preservation end of programme meeting in Birmingham. I recall being very nervous at my first JISC event in November 2007 but feel much more at ease now and enjoyed the event immensely; the programme has certainly been successful in fostering a sense of community though it’s an unusual social experience to meet people face to face, often for the very first time, when one feels you already know them from reading their blog and following them on Twitter.

During one of the breakout sessions on the first day I made a bee-line for a discussion about repositories for learning and teaching materials – as opposed to OA research repositories. I use the word “opposed” advisedly as there is certainly some strong sentiment around the issue, particularly with respect to using a common software platform. As a representative of a project that is adapting a learning object repository to also serve as an effective Open Access research repository I’m finding it a little difficult to understand the vehemence of some of this opposition, though I would be the first to acknowledge a steep learning curve and recognise that we have required extensive development, not of intraLibrary itself perhaps, but of an appropriate web infrastructure surrounding it. And yes, we would certainly have been able to implement a functioning OA research repository more quickly using EPrints or DSpace however, from the outset, it was vital that our repository had the the capacity to fulfil its broader potential – in the words of Clifford Lynch “[A] mature and fully realised institutional repository will contain the intellectual works of faculty and students – both research and teaching materials – and also documentation of the activities of the institution itself in the form of records of events and performance and of the ongoing intellectual life of the institution.”  [Lynch, Clifford. A “Institutional Repositories: Essential Infrastructure for Scholarship in the Digital AgeARL Bimonthly Report 226 (2003).]

It’s also important to be pragmatic.  Historically, Leeds Metropolitan University is a polytechnic that gained chartered university status in 1992; its heritage is very much in teaching and learning rather than research with, arguably, a more vocational than academic flavour.  In recent years, the research profile has steadily increased, culminating in unprecedented success in the 2008 RAE and the university is naturally keen to capitalise on this success, enhance its research profile further whilst also continuing to emphasise its student focussed teaching and learning credentials. The implementation of an integrated repository to support both research outputs and learning objects reflects this dual focus.  Clifford Lynch’s article suggests that the concept of a central system to manage disparate resources in this way has been implicit within the sector for some years, however, the technology has tended to focus on Open Access to research, with the two most widely used software platforms being EPrints, developed at the University of Southampton in 2000, and DSpace, developed at MIT in 2002; early versions of both platforms were primarily designed to manage text based resources (though subsequent versions of EPrints and DSpace can manage a wide range of digital file formats.)  

NB.  In an extended discussion on this issue on JISC-REPOSITORIES (archive hereRepositoryMan Les Carr of EPrints refers to the fact that he still comes across the firmly held (and spurious) belief that because EPrints is used for Open Access it can’t be used for multimedia files or scientific data.

The session was chaired by Amber Thomas of JISC and I asked a somewhat blunt, perhaps naive, question about JISC’s perspectives on combined repositories of research and teaching materials.  Amber suggested that JISC have been deliberately neutral on the issue which is also perhaps emphasised by the diagramatic representation of the programme structure reproduced above.  

Some of the commentators last Wednesday were adamant that though it may well be possible to manage different types of resources with a single system it was far from desirable with one colleague making the pithy analogous observation that you can write letters in Excel but that doesn’t make it right.  Phil Barker of CETIS was also at the discussion and in a recent blog post on the “question of whether research outputs and learning materials should stored in the same repository” is “inclined to think the answer is no, the purpose of the repository is different, a learning material isn’t an output, sharing means something different for the two resource types.”  Phil goes on to say that ” If you think a repository is a database and a bit-store then you may come to a different conclusion, but I think a repository is a service offered to people and your choice of starting point in offering that service will affect how easy your journey is.”  (Full post here)

I’d certainly concede that our journey hasn’t been an easy one and I also agree that a repository is a service offered to people and with our repository start-up, and also Streamline and PERSoNA, that is certainly the approach we have tried to take; with intraLibrary and the SRU interface we now have an incipient infrastructure to manage both research material and learning objects; the discrete types of material can be managed entirely separately, however, there is also potential for the ongoing development of a holistic approach to the management of the full range of digital resources produced by a modern university and as we develop our infrastructure further I hope we can utilise appropriate web-technology around a central management system (intraLibrary) to achieve decentralised resource discovery – through appropriate interfaces, widgets and environments – the VLE for example.


Then of course there is the small matter of persuading academics to part with their resources, not to mention IPR, copyright and quality control issues…

Open Access to research is an evolving paradigm and represents a considerable shift in the established academic publishing process; Open Access to a broader range of educational resources still more so. Any paradigm shift is likely to take time to evolve and Open Access, to research and other materials, is no exception, especially given that academia, perhaps, tends to subscribe rather strongly to established tradition!

JISC’s current OER programme should go some way to addressing many of these issues but infrastructure is the foundation. The perfect system almost certainly doesn’t exist and it’s surely important to be pragmatic when implementing and developing appropriate system. Here’s to ongoing discussion, debate and development.


intraLibrary is just for Learning Objects, isn’t it?

The issues around adapting intraLibrary to adequately function as an Open Access repository of research are agonisingly documented on this blog (see; there was an interesting and necessary discussion on the JISC-REPOSITORIES mailing list yesterday (though it’s still rumbling on) about the differences, if any, between ‘general repositories’ like EPrints and DSpace and specialised Learning Object repositories and the suitability of various platforms to fulfil a variety of institutional needs (Open Access to research material; Reusable Learning Objects/Content Packaging; other multimedia and complex digital objects. )

The hardest line was that it would be highly impractical to use the likes of EPrints and DSpace to “store, catalogue and serve e-learning resources” or, conversely, to use a specialised LO repository like Intralibrary for research.  My own view is that there is scope for complementary technology and that LO repositories can benefit from the culture of openness and sharing exemplified by OA archives of research as the zeitgeist shifts towards Open Access to a wider range of educational resources.

I would be the first to recognise that intraLibrary isn’t ideally suited to be used as an Open Access repository, however, with some “customisation” it can do the job perfectly well.  I expect the same is also true, from the other direction perhaps, of DSpace and EPrints – Soton, in fact, is currently  developing EdSpace based on its famous open source software.

Institutions increasingly expect their repositories to manage a wide range of digital material; at a recent RSP focus group it was clear that repository administrators running a range of platforms are increasingly being expected to manage everything and the kitchen sink.  Moreover, institutions, especially smaller ones, simply don’t have the resources to implement the ideal software solution(s) that will satisy multiple stakeholders.

(Disclaimer:  Some of these perspectives are paraphrased from my colleagues on JISC-REPOSITORIES)

For Les Carr’s perspective on the discussion* see his blog post on repositoryman.

* or argument!

Software announcement

I can finally announce that intraLibrary from Intrallect has been selected as the software platform for LeedsMet Repository.

Originally designed as a Learning Object repository, intraLibrary is the platform behind JORUM and will need some repurposing to also serve as an Open Access research archive. During our rigorous selection process and after careful liaison with Intrallect, however, we have been satisfied that such repurposing is achievable and that the software will ultimately provide the best all round solution for our requirements. We now join Oxford Brookes University’s CIRCLE project in using this software to implement a single repository for research outputs and Learning Objects.

intraLibrary will be implemented and configured over the next few months and I intend to start uploading research material almost immediately. An official launch, however, is still some way off while the necessary customisation is carried out. For demonstration purposes, priorities will be:

  • Development of appropriate workflows for ingest of research materials.
  • Integration of an SRU interface to facilitate open search and retrieval of research content.
  • Work with Intrallect to incorporate embargo functionality in line with publisher restrictions.
  • Work with Intrallect to incorporate report functionality (number of hits/downloads etc) that can be used in advocacy to the university community.

As initially prioritised in the project plan and due to the considerable amount of customisation to be undertaken, our early emphasis will be on research outputs; appropriate liaison will also continue within the university regarding LeedsMet repository and Learning Objects.

Which repository?

I recently posted under this title on the Streamline blog and it has raised one or two issues in the consultancy phase of our institutional repository project that I wish to explore further in an appropriate forum i.e this blog. Although this is primarily of interest to the respective project staff at Leeds Met, I am, of course, also interested in the opinions of others external to the organisation.

I have been grateful for the input from the Streamline project team and cross project work is obviously extremely valuable in that it brings together different perspectives; in this case, precisely what we require from our repository software.

The start-up phase of the (institutional repository) project will focus on an Open Access research archive, however it is extremely important that whatever solution we opt for is extensible to other types of repository content (Streamline, for example, is looking at the work flow associated with the use of learning object repositories and is developing a suite of tools and practices that will reduce the administrative impact of this on teaching and research staff.) Moreover, looking towards the future there is a broader issue that we do not commit ourselves unnecessarily to a plethora of discrete information silos that become impractical and expensive to integrate.

As I have discussed with my colleagues, in terms of the specialised functionality required by an Open Access research archive, personally I was swayed by the demonstration from (drum-roll) Eprints and, to a lesser extent, bepress’ Digital Commons. There have been concerns raised about Eprints’ (and others’) search interface and while I’m not certain that I necessarily share these concerns I presume the interface could, in any case, be heavily customised to our needs.

NB. A colleague at a recent networking event suggested that there is very little evidence that people actually search institutional repositories at all and tend to come to an IR through Google or another search engine (I realise this is hardly an argument in favour of a flawed search interface but it does illustrate how an IRs priorities may differ from a those of a more general repository).

It could certainly be argued that IntraLibrary, DigiTool and HarvestRoad Hive would represent a more holistic repository solution for the University as a whole and (obviously) the respective costs of each platform will also need to be taken into account. The other crucial factor is time-scale with the (start-up) project funded until March 2009 by which time we are commited to having the (institutional) repository embedded in the work flows of the respective departments of the University and to having a “representative body of content”. Pragmatically this means we need a functioning repository as soon as possible so that I can move on to what will be the meat of my role; promoting the IR and the principles of OA to the research community and getting the repository populated – we know from others’ experience in the sector that this is likely to be the biggest challenge of all.

Right now however, the burning question (for me at least) is “Which repository?”