We have lift-off! What next?

All in all I thought the launch party on Friday went really well, apart from my lacklustre attempts at carving an enormous, JISC branded (really!) cake with a very small knife. And where did all those students come from?  Pre-cake speeches from Barbara Colledge and Wendy Luker, then, were to a sizable crowd who applauded dutifully when Shiela Scraton, Director of University Research cut the golden ribbon, though we still had a good turnout even after the enormous confection and the students had disappeared (presumably back to their desks to scribble up essays in a frenzied sugar-rush.)

I would like to publicly thank Dominic Tate from The Repositories Support Project who, after the ceremonials, gave an excellent presentation introducing his Friday afternoon audience to the growing global infrastructure of IRs of which we are very-nearly a part (launch party notwithstanding we are not quite ready to register with OpenDOAR, more on that in a minute) and covering some of the key benefits of OA to research. I followed Dominic with a demonstration of our repository – both the Open Search interface and intraLibrary itself – emphasising that we are also managing RLOs as well as research, and hopefully gave the audience some idea of where we hope to be by the end of the project in March.

NB. A peculiar and occasional machine specific bug (i.e. it only seems to happen on certain PCs on the LeedsMet network) meant that I couldn’t actually access the Open Search interface on the development server which was frustrating and had to use my demo video – I only belatedly remembered that IMTS had made the interface live on the production server for me – http://repository.leedsmet.ac.uk/main/index.php – however, I did encounter the same problem with the new URL on a colleagues PC yesterday so if anyone has difficulty accessing the interface please let me know. When the problem happened with the dev server IMTS were baffled.

So…what next?  We may have successfully lifted off from the gantry but we are still a long way from achieving a stable orbit and with a little over 3 months before the end of the project a rough list of our priorities might look like this:

1.  Content: research

• I’m reasonably happy with the metadata template we are using for research and the way that metadata is being returned and formatted in the Open Search interface.  Both will be subject to comment by a wider cohort of staff over the coming months and no doubt extensively tweaked.

• We do perhaps need to think carefully about the ‘keyword’ field that is being mapped onto DC Subject.  We almost certainly want multiple (uncontrolled?) keywords or do we wish to use a controlled vocabulary?  Which one?  Or do we just want (multiple?) uncontrolled keywords and classify more formally against LOC?

• I hope to start training a couple of colleagues from the library very soon and they can start uploading research material – before Christmas I hope – their input should also inform the questions above as well as broader metadata considerations and workflow development (see below)

2.  Content: Learning Objects

We have identified several potential sources of content including:

• Resources already in X-Stream (the VLE) and it’s associated storage space

• Skills for Learning & Centre for the Built Environment – both of whom have web-based resources to catalogue by URL

• Learning Technologists, being faculty based, will have access to a range of LOs and should probably be the first user group to set up with intraLibrary accounts and let loose in the repository.

Progress will be subject to establishing viable workflows and appropriate metadata templates, concomitant user-testing and work being undertaken by Streamline.

3.  Workflows

The implications for workflow are naturally going to be different depending on type of resource:

• For research, in the first instance, the workflow will be fully mediated by myself and colleagues from the library.  As mentioned above I hope to start training colleagues very soon and I hope the workflow will evolve organically though we may well have staffing and resourcing issues depending on the amount of content identified and submitted – this, in turn, will depend on the effectiveness of my advocacy work of course!

• In the longer term we are hoping to implement a quick-deposit facility using SWORD, there will be implications here for authentication (see below).  Moreover, what will happen to resources deposited in this way?  Presumably they will have to go into a (library) mediated workflow meaning further staffing and resourcing issues.  What metadata will we be able to generate from deposit?  At the very least we would want depositor ID.

NB.  This is unlikely to be realised before the end of the start-up phase of the project.

• For Learning Objects the workflow will be entirely different and, at this stage, I am much less clear how it might work though anticipate it being relatively straightforward as intraLibrary, after all, is a purpose built LO repository.  The work that is being done by the Streamline project will be important here and I am liaising closely with Dawn whose recent blog posts begin to tackle these issues here and here.

4.  Authentication

Once again, the implications for authentication, I think, are somewhat different depending upon type of resource:

• For Learning Objects pragmatically I think we need to be able to provide ALL staff with some sort of access to intraLibrary. There will naturally need to be different levels of user with some having browse/search rights only; others being able to upload and yet others having full admin rights – I expect I can manage such a hierarchy through intraLibrary’s internal group management and it wouldn’t necessarily have any implications for a standard authenticated account for all (N.B. Might we also need to give students access to intraLibrary?)  Currently I am manually setting up user accounts with standard network usernames, but will this mean that when authentication by LDAP is implemented then these accounts will be duplicated and that any resources associated with the ‘old’ accounts will need to be moved/uploaded again?

• Authentication is perhaps less of an issue for research material as it will be accessed on an Open Access basis via http://repository.leedsmet.ac.uk/main/index.php , however, currently we require authenticated access for library/admin staff to upload resources – again, I am currently achieving this by setting up manual accounts – will these be transferable to authenticated accounts?  In the longer term we are hoping to implement a quick-deposit facility using SWORD which I imagine sitting behind a Leeds Met authentication barrier (like the portal?) and that allows research staff to browse for a file on their hard-drive and upload it to the repository (to a librarian’s workflow?)  In this scenario we would want to know who had uploaded a research paper and I’m not sure what the implications of this might be in terms of capturing log-on information and having this recognised/populated in intraLibrary.

5.  Policies/usage agreements/licencing

• These will be more straightforward in the case of research material where the main issue is around self-archiving permissions according to publishers’ and individual journals’ copyright transfer agreements (SHERPA/RoMEO).  However, there will be implications communicating the issues to research staff and for work-flow (i.e. the necessity for an academic, in most instances, to provide their own final draft of a paper for upload.)

• The issue is potentially much more complicated for Learning Objects and will certainly require further consultation and input from the University community.  Dawn’s perspective, and I tend to agree, is that, initially, ALL LO’s should be restricted to only being discoverable to authenticated users within intraLibrary before we start worrying about making any of it openly available – though perhaps we could make metadata publicly available.  See Dawn’s post here.

5.  Search Engine Optimisation/OAI-PMH/registration with appropriate services

• Ideally I would like to facilitate full text indexing by search engine bots but I’m not certain if this is currently possible with intraLibrary and we may need to look at setting up Google site-maps in the first instance.

• Though I don’t think OAI-PMH is supported by Google any longer, it will still be important for third party harvesting services like OAIster.

• I will need to review what other services we should register with.  The obvious ones, as Dominic mentioned, are OpenDOAR and ROAR.  We do not yet have enough content nor are we slick enough to register but I am confident that we will be by the end of March!

6.  Developing infrastructure

• PERSoNA

We are making good progress with the PERSoNA project and the main output will be a website comprising a variety of tools, widgets and links to facilitate social interraction with the repository.  A very preliminary blog has been set up at http://leedsmetrep.wordpress.com/ .  As it is hosted by WordPress it is difficult to add widgets to at the moment so the first job is to set it up on our own server space.

NB.  There are some exciting tools being developed that will tie in with PERSoNA like Stuart Lewis’ Facebook SWORD app – http://fb.swordapp.org/

• PowerLink to X-Stream

This is already installed on the X-Stream test server and is kind of working albeit with a few bugs.  It would be nice if we can get it working properly for March but probably not an essential output at this time.  We will also want to enhance its functionality but that is almost certainly for a future project.

So that’s it really.  Not much to do.  And I never did have a piece of cake!

Repository Steering Group meeting: 22nd July 2008

The staff development festival in September is a unique opportunity to promote the repository and our agenda for yesterday’s meeting aimed to get some much needed input from the steering group before the quiet month of August.

Item 1. Recap of previous meetings:

Documentation approved.

Item 2. Update on progress with intraLibrary

2a. Configuration:

Search interface (SRU):

Getting the search interface on line is the first priority – my request for the server is still pending with IMTS but I hope we can install the IRISS interface as is within the next few weeks (JohnG is installing it on a local server as we speak which can then be tranferred to our Leeds Met domain when it is available) and I think it will be straightforward to switch the CSS to get a very rough Leeds Met branding.

Content structure:

This is also crucial and needs to be put in place ASAP. Several members of the group expressed the opinion that it should not be based on faculties which tend not to be fixed entities within the university; it was also thought that such a schema would not reflect institutional emphasis upon cross-disciplinary research. There was consensus that organisation at the top level should be by content type (i.e. Research/Learning Objects) but exactly what hierarchy should be employed beneath is still not clear (library of congress subject headings?). We also need to make a decision on what other material types will be accomodated in the prototype (e.g. Dissertations and Theses)

Landing screen:

Technical challenges aside, the current conception of the landing screen is that it will essentially use the same template as the search interface i.e. it will be branded the same and share the same look and feel; it will also share some of the same functionality and link back ‘home’ to the search interface.

Given the close relationship between these configuration issues, a sub-group was identified that will liaise as necessary to develop the content structure; branding; look and feel; usability and will also inform the technical development of the additional functionality.

2b. Policies:

The group was briefed on the types of policies that need to be developed (see last post) with emphasis on the fact that the ‘standard’ institutional repository policies may be insufficient for our requirements given our wider remit (i.e. not just research outputs). A sub-group was identified that will liaise as necessary to develop suitable policies.

2c. URL:

The suggestion mooted – repository.leedsmet.ac.uk – was deemed suitable by the group

Item 3. Content for the repository:

To discuss method of contacting researchers / research active staff and soliciting content

Review of draft correspondence for research active staff and discussion of when this would most usefully be disseminated; consensus that it would have the greatest impact some time after the staff development festival. Content was broadly approved though it was suggested that greater emphasis be placed on the benefits of OA to citation and the increased importance of citation under proposals for REF (to replace RAE).

Emphasis was placed on the need to identify and recruit interested parties within specific faculties/research groups to help drive the advocacy process to the wider community; liaison with University Research Office for appropriate contact lists.

(NB. This is an ongoing process that is already underway but will increase in profile with the implementation of the prototype system.)

The Staff development festival confirmed as a key opportunity.

There was discussion whether content would be full text only or would also comprise citation of material that we do not have copyright permission to make available as full text (i.e. bibliographic reference only). Given that including such material will enable us to ‘hit the ground running’ and considering the increasing importance of citation data/bibliometrics for the RAE / REF the consensus was that citations should be included at the outset.

Item 4. Authentication

It was emphasised to the group that we can be fully functional as a mediated repository without the need for authentication in the first instance.

A representative from IMTS was able to inform the discussion in the light of recent feedback from Intrallect and will continue to liaise as necessary.

Item 5. Integration with other Leeds Met systems

In light of the decision to include citations as well as full text, an important early integration will be with SFX such that citations in the repository can incorporate a link to Leeds Met holdings of subscribed material; hardly Open Access as it will only be available to authenticated staff and students but will offer another local route to that material and can also be used to generate data on OA friendly publishers and perhaps to raise awareness of OA.

The PowerLink to X-stream should also be a priority such that it is operational at the earliest opportunity.

NB. Precise functionality of the PowerLink still needs to be determined.

Other systems flagged up for integration were iTunesU and the streaming server; pending investigation!

The next meeting of the steering group will take place after the staff development festival, probably late September/early October.

Search interface, URLs, taxonomy, policies and content…

It is now established that we will be using the SRU interface developed by IRISS as the public search interface for the repository. I hope to install the current incarnation of the interface on a Leeds Met server very soon and two of my more technically adept colleagues are looking at the recently released code in order to scope the extent of the development work that will be required to incorporate advanced search and browse functionality. As this page will effectively be the repository by proxy (the URL that I have requested is repository.leedsmet.ac.uk – intraLibrary itself will require a different URL) we also need to think about what other elements it might need to comprise; authenticated log-in to intraLibrary itself (yet to be determined if this will be the appropriate route for self-archiving; it will certainly be one route but we may also need an authenticated link to a SWORD interface for example); About this repository; FAQs; Operational policies; Contact etc. It is also likely that this page will form the basis of – or at least link to – the PERSoNA web-tool(s).

What about learning objects which will require their own taxonomy and a different workflow for deposit (via SWORD perhaps)? Should they be incorporated into the search interface at all or will users need to authenticate into intraLibrary to browse? This would seem to make sense given intraLibrary is a specialised LO repository and access to this type of content is more likely to be restricted to Leeds Met staff.

I’ve adapted my schematic recently posted on PERSoNA News to try to represent what the repository might now look like:

The customisation of the search interface is one of the issues that I am taking to the steering group meeting tomorrow afternoon.

Other decision that needs to be ratified by that group are:

• The URL for the search interface
• The URL for intraLibrary
• The taxonomy system that we shall use within intraLibrary and that the search interface (SRU) will map directly on to (at least for research)

Other items on the agenda are:

• Development of operational policies for the repository

I have so far drafted the following:

1. Metadata policy
2. Data policy
3. Takedown policy
4. Content policy
5. Submission policy
6. Preservation policy

These are all fairly standard in terms of Open Access repositories and, with the exception of 3. Takedown policy, were all generated using the OpenDOAR Policies Tool, nevertheless, it may be necessary to identify specialised sub-groups to review these drafts to ensure they are appropriate for the Leeds Met repository; the issue is more complex of course due to our repository incorporating Learning Objects as well as research.

• Content for the repository

There needs to be a discussion about how best to contact researchers and research ac tive staff to ask them for appropriate material for the repository. In the first instance, in line with the project plan, this will be their own versions of published research articles that are allowed to be self-archived into an OA repository. I have begun to identify such material and have drafted correspondence for review at the meeting.

• Authentication

With the implementation of the search interface (SRU) it will not be necessary to authenticate in order to browse for research content (essential for OA). It will, however, be necessary to generate authenticated accounts for Leeds Met staff that require access to intraLibrary itself and these will need to be integrated with LDAP. Though much will depend on the precise configuration of our integrated repository systems it is likely that, in time, all staff will require an authenticated account whether to deposit material, search for learning objects or access their internal workspace. There are also authentication issues pertaining to the potential use of SWORD/other external interfaces such that only authorised Leeds Met staff/students can deposit material/access federated content. I am still unsure of some of the issues involved and require input from Intrallect and IMTS.

• Integration with other Leeds Met systems

This is an area where it is perhaps still too early to think much beyond priorities and broad timescales. Given that there is already a plug-in for X-stream and that this is functionality that can be used as a selling point to the university community it makes sense to focus on this integration first. Also, perhaps, library online and the portal.