cgspace / ILRI / ILRIComms / Knowledge and Information / Open access

From mahider to CGSpace: Moving our research outputs to open repositories

In late 2009, the International Livestock Research Institute (ILRI) began work on an institutional repository using Dspace software.  Under the name ‘mahider’ (, we intended this repository to be a complete index of all ILRI research products/outputs; with ‘full’ content as often as possible (we struggle to get full versions of articles and other products published by commercial third parties); and with as wide a coverage as possible.

The repository today has more than 6000 entries with very good coverage of our 2009-2011 work, as well as some older archive materials.  Excepting roughly 1000 journal articles, the remainder of the repository contains or links to fully open resources. Since January 2011, we have been publishing these under a creative commons license.

As we progressed with our repository, we discovered that a lot of our outputs were not wholly ‘ILRI’ in origin. They came through wider projects like East Africa Dairy led by Heifer, the CGIAR Systemwide Livestock Programme (hosted at ILRI), or through the Nile Basin Development Challenge (with the International Water Management Institute). Moreover, we were also capturing all the materials associated with the emerging CGIAR ‘megaprogram’ on livestock and fish – a joint venture with 3 other CGIAR centers. We could not put all of these under a purely ‘ILRI’ banner.

Talking to Bram Luyten at the @mire company in Belgium, we discovered we could use our Dspace installation in a multi-inhabitant mode – ie, we could present the various different groups independently on the same platform.  These ideas were confirmed in a Dspace training session run by @mire in Nairobi at the end of 2010. Here we were joined by colleagues from the CGIAR Challenge Program on Water and Food (CPWF) and the World Agroforestry Center (ICRAF) with the idea that we could collaborate and share the same Dspace installation. Importantly, Dspace allowed for each community to have its own URL, look and feel, and administration.

During 2011 therefore, with the CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS) joining us, we extended the orignal ‘ILRI’ installation into a shared platform that we called ‘CGSpace‘ – similar to other CGIAR-wide initiatives.

We devised different themes for each group as well as templates and other specific vocabularies.  We began to migrate and add content from existing systems. We upgraded our Linux server and server management. This was a steep learning process and we learned a lot about the need for quality checking and standardization before importing. We also learned a lot about shared administration and server management, the need for good documentation and clear updating routines (we now have both live and test servers and our code changed are deposited in GitHub)

While we were working on our Dspace, colleagues in other CGIAR centers were also setting up their own installations (CIMMYT, ICRISAT, IWMI) and others were interested to perhaps join with us (CIFOR, ICARDA, WorldFish). One strength of Dspace is that any installation can harvest the metadata from another.

Thus, CGSpace can pull in and display metadata from other installations, at IWMI or CIMMYT for example. These centers can run their own installations and be part of the wider CGIAR-wide effort. For a center that does not want to run its own Dspace, they can join the shared installation and share the hosting and support costs. We are looking at both variations and continue discussing across several CGIAR centers using a Yammer discussion space.

In November 2011, the CGIAR Consortium indicated they wanted to formally join the project to publish their archive and new content on the platform. They also indicated that Dspace would be one of ‘the’ collaboration tools supported across the consortium. Good news!  It also takes the project to a new level that will be the main focus of this work in 2012.

As an important aside to this Dspace story, we agreed with our data management colleagues in the ICRAF-ILRI Research Methods Group that our Dspace should also index and point to published ‘data’ from our centers.  While we concentrated most on Dspace, our ICRAF colleagues investigated options to use a similar system for data sharing and management – Dataverse (which is also used at IFPRI). The results look positive (see presentation by Anja Gassner), however we still need to make the concrete/precise bridges between our Dspace and Dataverse worlds.

The whole issue of more available and accessible data and research outputs has been on the CGIAR agenda for some time, and is increasingly important in the new CGIAR research programs which call on centers to work much more closely with each other and other partners, learning, sharing, and communicating as they do so.  Most important, the new performance agreements call on us to make our outputs and data open and accessible to the world.  Dspace and Dataverse are some of the ways we think we can do this …

2 thoughts on “From mahider to CGSpace: Moving our research outputs to open repositories

  1. Peter and the Team, thanks for sharing this learning experience. What I appreciate the most about the post is not necessarily the decision to adopt a common tool itself, but the process of concertation (french influence creeping in here, read ‘consultation’ instead) among peers, the realisation that a single hosting of the tool allows for various interfaces or front-ends for the various users, and finally the value-adding that came with such a decision, with some room for future development with Dataverse.
    NARS institutions are often faced with similar challenges of which platforms to adopt, how to host these platforms in a sustainable way, whether there is a way of sharing the costs and more importantly whether there is a way of ‘learning by doing together’ so we live the experience together and support each other as we develop our repositories.
    Your experience demonstrates what many NARIs could also be adopting as an approach, through a shared hosting arrangement. Now this could be done at a national level or even at a sub-regional or regional level. In the end, what matters is that the information resources we possess are more accessible, and we can still present ‘our institutional collection’ as a subset of a larger collection.
    What I would now request is that your colleagues of the CGIAR document their various implementation and importation scenarios and the lessons learnt in the process, because these will be invaluable information for the isolated ‘information manager’ in a NARI to visualise the steps that will have to be taken to get on board such a shared platform. Personally, I would love to be able to read about your experiences and re-live them before I embark on such a journey or initiate a collective action in this direction.
    Thanks again for an inspiring post to start the year 2012!
    Krishan Bheenick
    Food and Agricultural Research Council, Mauritius

  2. Peter et al: Congratulations, this is an excellent move forward that should improve the access to and use of these valuable resources and, at the same time, help bridge the gaps that often separate disciplines and their contributions to solving R4D challenges.

    It will be interesting to see how the development of CGSpace can underpin the CRPs and their information sharing and how it can serve to strengthen their linkages to national, regional and international partners and to the end-users of the research.

    Bill Thorpe, Cambridge, UK

Leave a Reply

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

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

Facebook photo

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

Connecting to %s