Leeds Met Repository Open Search Version 2.0

This is a bit of a trailer for our shiny new interface that I hope will go live in the next week or so and a run down of some of the new features.

It’s far from perfect and should still be seen as a beta – we very much need real users to start using it and I’m feeling a little nervous about how it will be received as I know how much work Mike, in particular, has put into it.

The interface has evolved from an SRU client developed for by IRISS – http://www.iriss.org.uk/learnx – which is available under GNU General Public Licence v.3 at http://code.google.com/p/sruopensearch/ (N.B.  We still intend to release our modified code under a similar licence.)  Learning Exchange Open Search is a great front end for searching intraLibrary but with just a simple search box lacked advanced search functionality that was essential for us.  We also wanted to use intraLibrary to manage resources for teaching & learning aswell as facilitating Open Access to our research collection in accordance with the EPrints model.

The tabbed interface incorporates an “Advanced search” form that allows users to cross reference multiple fields specifying AND/OR and they are also able to search for either “Research” or “Open Educational Resources” which uses authentication tokens to return results from the appropriate collections in intraLibrary:

There are also big changes in the way that results are returned; Mike has been able to use a unique identifier to build individual pages for each record so that a search will return a set of results that indicates whether or not each individual record has the full text available:

These titles then link through to a static HTML page comprising all of the metadata associated with that record including a published URL and, where the full text is available, a link to the PDF in intraLibrary:

This static page should be indexed more effectively than was the case before though there is one small fly left in the ointment in that the public URL generated by intraLibrary that is used to download the full text is dynamic which means it cannot be indexed by Google; I’m not sure if it will be possible for Intrallect to do anything about this though they are aware of the need for full text indexing and are looking into the problem.

Spinning plates: A repository update

I feel at the moment that I’m trying to spin plates, rushing around like a guest act on the Paul Daniels magic show to impart sustained rotation – when I get to the last plate the first is beginning to wobble….and then someone throws me another, just to make it more exciting.

Oh and all the plates are different shapes and sizes (some of them I swear aren’t even symmetrical which gives them a lop-sided gyration that is really tricky to maintain…)

Currently research material collectively comprises one of the biggest and wonkiest sets of crockery:

1. Peer reviewed journal articles
2. Book items
3. Conference proceedings
4. Conference items (e.g. ppt presentations)
5. Theses or dissertations
6. Reports

All of these need to be displayed differently by the Open Search interface which means Mike is currently hacking away at the code to ensure they are picked up by Educational properties: Type of resource and then formatted appropriately.  In turn this is having a knock on effect for metadata entry and workflow and I’m still not sure how it will all tie together.

We are also relying on Mike to develop advanced search as part of the interface; work is progressing well and we are now able to cross reference (toggling AND/OR) Title; Subject; Publisher; Description (abstract); DOI; Type and Format  – the interface also has a free text box.  Mike is currently trying to implement probably the most important field – search by author – and has identified a potential problem due to limits on the way we can differentiate contributor roles in the metadata – currently we are only able to query Dublin Core rather than the full LOM record.  N.B. Recent testing has indicated that this may not actually be the problem we originally thought and it may be possible to query contributor role=author after all.  Then we’d also like to incorporate browse by author – though this is currently another plate yet to be balanced and spun….

Then there is the ongoing issue of differentiating content (i.e. research material vs. learning objects) and ensuring these are returned appropriately by an extended/alternative Open Search interface – this functionality is crucial to Unicycle of course to make OERs available.  We’re currently exploring using collection tokens which should allow us to submit a query incorporating an authenticationToken such that a given query only returns content from a particular collection.  Initial testing has gone well and we’re pretty confident we can implement this over the next week or so.

(NB.  I need to start thinking about license models for Unicycle and some flavour of Creative Commons will need uploading to intraLibrary, another one for the to do list…)

Ideally we could do with a functioning PowerLink for the VLE by September – as well as being crucial for the PC3 project, it represents functionality that will really get people engaged and allow us to demonstrate the benefits of storing and sharing teaching and learning materials in the repository rather than within the modular, inaccessible silos of Blackboard-Vista.  Also important for PC3 is implementing LDAP authentication (which really should be happening soon!) thereby giving teaching-staff – and the first PC3 cohort – access to intraLibrary – NB.  PC3 doesn’t require Open Access in the same way as Unicycle.

I’ve also been working with Rachel on developing a workflow for CLA material and ensuring that we can generate suitable reports for the CLA – during this process, we uncovered a bug in the metadata editor which slowed us down a bit but with help from Intrallect, we’ve managed to implement a work-around pending the bug being fixed in a future build and Rachel has started using intraLibrary to store and disseminate CLA material on a pilot basis.

The most recent plate JISC have thrown our way, of course, is the Bibliosight project which will almost certainly have an impact on the developing infrastructure beyond the specific deliverables of that project – our first meeting is on Monday 13th July for which there is a draft agenda on the project blog.

I just hope that we can keep all the plates spinning and don’t end up with a Greek wedding scenario!