Aug 042011

Welcome to day two of Repository Fringe. We have opened with a little breakfast curtosy of the DevCSI hackathon and we have moved straight into our first section:

Laurian Williamson, RSP Project Co-ordinator for JISCrte programme – Introduction

Balvier Notay is also giving some additional background: in 2002 we had an exploratory phase of the programme – looking at OAI-PMH, then a building capacity phase starting up repositories, then an enhancement phase, then an a rapid innovation project – sneep and meprints came out of that, and then we had the Reposit project looking at workflows etc. they are just coming to an end, also various projects about automatic metadata also now coming to an end. The Repositories Take Up and Embedding programme was about getting these developments out there, about embedding repositories in institutions. We are creating a guide to embedding from RSP and a technical guide from Southampton.

Laurian is back to give some additional framing information. We have been encouraging take up of repositories, embedding, sharing good practice and experience. It’s a JISC funded project and we were working with six very different projects.

Jackie Wickham (RSP, University of Nottingham) – Overview of Repository Embedding and Integration Guide

We don’t have a guide to show you yet but we will publish it in September. How to embed the repository in the institution, case studies and video interviews are included. There will be a list of tools and apps trying to pull that together. There will also be a self-assessment tool to guage the level of embeddedness. This will be publicised very widely in the community when it is launched.

Xiaohong Gao (Middlesex University) – MIRAGE 2011 (developing an embedding visualization toolkit (for 3D images) and a plug-in for uploading queries)

This is mainly a medical images repository. We want to maximise the benefit for the community of this specialist repository. This project is running as part of WIDTH – Warehousing Images in the Digital Hospital – a project with 11 partners. We are also in touch with a German reposititory of a similar ilk, IRMA for bone age, and with a Swiss repository, HRCT, for lung images. Also a Greek semantic based repository i-SCoRe.

Our repository has been disseminated through three articles and a paper for eTELEMED conference won best paper as it explained the technical challenges of delivering 3D visualisations in the repository.

The enlarged project team includes 2 BioMed MSc students doing final dissertations based on MIRAGE database, also a PhD student working with us.

During this project we have enabled 3D visualisations and upload of images, our next stage is to digest 2D/3D movie data. We want to use Grid technology for this work. We also want to undertake some user evaluation and dissemination.

From the students point of view the repository has widened both expectations and experience. For the developers this has been a great experience.

Now handing over to Susan(?) for demo – there are over 2 million images in the database and you can search by image – the system looks at shape, texture, size, etc. A lot of medical images are 2D but can be combined into 3D or 4D images. A limitation for us was that you could see 3D images only as 2D slices. And you could not upload an image as a query image. We have added this via 3D Brain link – you can view the slices (and page through them) and view a 3D image alongside. And you can now upload an image from the internet to look for a comparison in the collection based on the image. You can also select relevant or non relevant results and research. Obviously this technique would also work for other types of repositories and searching and we are happy to talk more about this.

You can also view random images from collections as a browsing tool.

Marie-Therese Gramstadt (University of Creative Arts) – eNova (aims to extend the functionality of the MePrint profile tool)

    The reason we are doing this project is to improve the take up and embedding of repositories. The take up of repositories in the arts is really low so this is very important to us. This project is funded by JISC and run by University for the Creative Arts and University of the Arts London – we worked togetehr before on the Kultur project and we have been engaging with te Kulture2 group.

    We wanted to use the MePrints plugin – a profile tool for pages about depositors for EPrints. We have now installed the plugin to two repositories at the moment.

    How we have approached the project is to get feedback from the arts project. We have done work with 10 researchers on a long term basis (at least 3 points at which we are in contact). We are getting a lot more into the culture of the institution. The User Needs Analysis involved showing MePrints in use, looking at what researchers are using and a short survey (6 from UAL, 4 from UCA) – this will be written up as a full report soon.

    At the University for the Creative Arts the profiles are fairly textual. The University of the Arts London includes links to projects and materials in a more visual way.

    All staff had experience of staff research profile pages but most not very experienced with repositories. All wanted a web presence. We looked at some of their personal websites – they liked that there were no limits to their personal spaces online. We looked widely at what could or could not be included. We thought widely but then focused down.

    We created visualisations based on feedback and based on what was already in place. MePrints already has a space for a profile image but it would be great to be able to use video here. Rather than go crazy with social media we’ve focused on the key headings and use of controlled AHRC keywords for research interests – some require customisation to MePrints. The outputs tab includes categories of Publicaions, Exhibitions, Conferences – hopefully that would be from the repository but we hope to also add a field for material that can be linked to elsewhere – we think this may improve deposit as well. Finally the Gallery tab provides visual highlights of the depositor’s work.

    Alan Cope (De Montfort University) – EXPLORER (create workflows and processes to enable the embedding of the DMU repository within the DMU research systems and processes)

    DORA is the De Montfort repository. We had millions of items but a low self-depositing level and little connection to other research systems. EXPLORER had 2 strands: one to develop and implement workflows and processes to enhance and embed DORA within DMU research environment. Looked via focus groups and questionnaires but will also be looking at the Embed and Enrich projects. Strand 2 was to adapt and integrate toools to enlage DORA and enable deposit of a wider variety of outputs in line with REF2014. Looking at the Kultur, Air, MePrints and IRRA projects here.

    We had 81 survey respondents and conducted three focus groups. That was nearly twice as many responses as expected. Most respondents knew about DORA and most produced outputs as text but others create music, graphics, photos, etc. We asked what might make people use DORA more and they commented that the production of statistics and the reuse of data held in DORA – not having to resubmit multiple times. They also suggested ways to improve the process and look and feel.

    We will be creating an updated process map – previously each depatment had their own guide. We are simplifying it down to one process. We are improving and planning advocacy and be more proactive with that.

    The key technical work for strand 2 is to improve the display of non text items. We will create a Kultur plugin or DSpace (as Kultur doesn’t work in DSpace). We want images and video that works on th epage – video is now working, looking now at other media types.

    One big advantage is that the university is to have a new websites – we wanted to integrate DORA more closely with this and feed DORA outputs to new individual researcher profiles. And we are currently looking at the functionality from AIR, IRRA and MePrints. We have also been testing the CERIF4REF plugin. It works in DORA but need to review what is pulled out, test it, and see if it is right for us.

    The Survey was very successful and provided much information. The technical work has been more complicated than expected but will provide useful functionality for Dspace users on completion. And we are currently documenting the new processes. See:

    Miggie Picton (Northampton University) – NECTAR (implement new tools, procedures, and services to enhance their repository – in readiness for the needs of the REF and EThOS)

    This project came out of the fact that NECTAR has been there for a few years and had done well with collecting metadata we have not done so well with geting full text. We had a very mediated process – which our researchers were keen on. We had a bit of a nudge as theses mandate was about to come in, and the REF is of course also a key and timely driver.

    We wanted to modify the university procedues for submission to NECTAR and to make some technical and procedural changes as neccassary to ensure NECTOR connects to eTHoS. We also wanted to keep up with rebranding. And we wanted to bring in added value from the sector – ways to make it more valuable to repository users, particularly researchers. And we also wanted to provide training and advocacy an we also wanted to collaborate with colleagues at the RSP.

    So far we have rebranded NECTAR to match the current University’s website. EPrints services have implemented the Kultur extension onto the NECTAR test server – we have used this to get testing done before doing live. So we have added things like the scrolling display of images on the homepage and the reformatted item pages which present full content (if available) before metadata. We have also made some changes around the home page – we did have a top 5 papers. We mainly did this as they rarely changed much and that can be discouraging. Now it’s a latest additions list. And we’ve made some other minor changes – one user had problems with the search box so we have added a more obvious link to the advanced search.

    We are now working on making NECTAR ready to talk to EThOS. We are also working with the Graduate School to gather theses. We have only awarded research degrees since 2005 so we should be able to get a full metadate record for all theses and may be able to get full text from alumni. Procedures for depositing into NECTAR is now part of the research degrees hand books. We have also altered the data entry process so that researchers can enter their own data but it is still a mediated process and that’s what our research committee want.

    Ongoing advocacyhas involved presentations to research groups, school research forums and with school NECTAR administrators and academic librarians. NECTAR training is now in the universititys IT training programme. And we are having an Open Access week event to raise the profile.

    Researchers are now notified by email when items are deposited in live NECTOR (woth thanks to William Nixon). And the NECTAR bibliographies have been revised in response to researcher feedback about how to present publications etc. The University web teamn will be including NECTAR bibliographies on staff profile pages. We’d love more ideas like this!

    Still to come… Changes for the REF via an Eprints plugin, And the promotional campaign for our Open Access Week event.

    But on the wish list: we want improved statistics on usage and dissemination to users; integration of NECTAR with a university CRIS – and our research office would be all for that. And I’d like more staff for NECTAR!

    Chris Awre (Univ of Hull) – implement the Hydrangea software

      I will be talking about hydra in Hull. We are a collaborative project between University of Hull, University of Virginia, Stanford Universitym, Fedora Commons/DuraSpace, MediaShelf LLC. This was an unfunded project for a reusable framework for multipurpose multifunction munti-institutional repository-enabled solutions. The solution is modeled to be useful to our own and others repositories. We initially set ourselves a three year time frame from 2008-11 but we have agreed to work together indefinitely.

      The Hydra project was funded to apply these solutions to the University of Hull website. We are working with MediaShelf as a technology partner and their role as part of this project is to do a lot of the implementation and do a lot of knowledge transfer. I am pleased that our developer and a colleague have picked up and loved this work so we know we’ll be in a good place moving forward.

      The project had three phases – a read only interface; ingest and metatdata edit functionality launched in June; full CRUS capability for some genres should be done by September and we hope to replace our interface by then as well.

      The idea is to create a flexible interface over a repositoryy that allows for the management of different types of content in the same repository – the end user has a single place for all materials and we think that supports embedding. And we think this encourages take up through our flexible development of end user interfaces where these are designed for the users according to content types – and separate management interfaces for repository staff. Hydra provides a frameowrk to support adaptability.

      There are  key capabilities: supports any kind of record or metadata; object specific behaviours (e.g. books, images, music, video etc); tailored views for domain or discipline-specific materials; easy to augment and adapt over time.

      We have developed partnerships from the outset as we think that’s crucial to the sustainability of the project. We don’t want too formal an agreement here but partners must feel comfortable with expectations. sharing of code etc. We are trying to establish a sustainable community around Hydra. For Hull specifically we are providing a UK reference implementation, creating a local knowledge base for others to tap into, and a place to start building a UK or European community.

      Hydra have developed guidelines around the organisation and structure of content – though the guidelines have wider applicability. Hydra runs on Fedora 3.x with a range of additional technology

      Hydrangea was Hydra’s initial reference beta implementation. Now deprecated but played it’s part. But we have for all our code. And here I shall borrow from a Prezi from Open Repositories to explain the technologies. We use Blacklight as a next generation library interface but it is content aware and metadata agnostic, and it has a strong community around it.

      Why these technologies? Well we all use Fedora. Solr is very powerful. Blacklight was in use at Viginia and now at Stanord/JHU. And Ruby allows for very agile programming approaches.

      Hydra in Hull creates records that look fairly traditional, though we couldn’t resist including a QR code. Usability testing was good. Lots of advice on improvements as well. We only had a few people but they provided great feedback. See

      Robin Burgess (Glasgow School of Art) – Enhancement of the design of their interface

      Glasgow School of Art have no public facing repository yet, they just have an inward facing tool so this is quite an important project for them.

      You can see I have a guitar. Being the Edinburgh Fringe I thought I’d give my presentation in the form of song!

      So we embarked on a project with JISC, we were very new to it but we weren’t afraid as we’d done our research… what we had found was that there was plenty of help around. We are building a repository, something new to GSA. Still don’t have a name. We looked at different systems – DPrints, PURE, EPrints, all quite similar. We decided to invest in DPrints.

      Requirements building is the next step, with some help from EPrints and Kultivate. We hope to develop an integrated system to help showcase the work at GSA. We have to move from Filemaker pro to something new with amuch better interface. We are building RADA, our new repository.


      Laurien adds a pre Q&A note that it is so important to disseminate this work and we are so keen to do this!


      Q1 – Peter Murray Rust) How many of your repositories can be exported as RDF

      A1) A few hands raised – from Northumbria, Hull, eNova

      Q2 – Les Carr) As a software engineer, if it were in the gift of people who build repositories to do one thing to make life easier what would you ask for? What have been your biggest problems?

      A2 – Northumbia) Magically change copyright law so that everything can be uploaded.

      A2 – eNova) It’s not really been about the technology, it’s the culture that is such a challenge..

      Q2) So Open Access Rohipnol

      A2 – Hull) Easily tweakable interface that can be tweaked according to need. Ability to change things easily.


       August 4, 2011  Posted by at 8:50 am Live Blog Tagged with: ,

      Sorry, the comment form is closed at this time.