The other panelists (including me) are now covering work at their instiutions (so excuse any sudden cut off as I have to speak!)

Firstly, Peter van Boheemen (Wageningen University, NL). They use Blackboard and Sharepoint. They have developed a function to allow the building of ‘course libraries’ in the library system. When the tutor links from the course to the library system, they are automatically logged in as the owning ‘course’ rather than a personal login. This sounds a bit like the work done at Westminster a few years ago, where MetaLib e-shelves were setup for each course…

Secondly, Kingsley Boulton (York, UK). They are using Sentient ‘Discover’ product for reading lists, and Blackboard for a VLE (apparently the Sentient product is now being acquired by another company, and will be called just ‘Discover’). They are only adding reading lists for those courses that have a Blackboard presence. They have a ‘e-learning librarian’ to do this work. They went live in January this year, and used a Blackboard ‘building block’ (developed by Sentient) to integrate. The reading lists are completely controlled by the academics. The idea is that library staff will rely very much on the lists in Discover to get details of what academics need purchasing, moving to short loan etc. User feedback at York has been generally positive. York are also looking at a building block to integrate MetaLib with Blackboard using a ‘Building block’ and the x-server. The aim is to have single sign-on between Blackboard and MetaLib, and be able to present specific MetaLib resources based on the student enrolments.

Thirdly, Bob Gerrity (Boston College, USA) is going to talk about using DigiTool with the University environment. At Boston College there is a IDETS (Instructional Design/e-Teaching Services) who work along side library staff to deliver services for e-learning. Bob’s got an interesting list of strengths for the library team, and IDETS team – something they try to keep in mind as they work together.

Boston College have already done hooks to library resources (from WebCT) – search forms, deep linking, links to subject specialists – this sounds very similar to what we have done at RHUL so far.

However, they are now starting to use DigiTool to provide a platform for faculty members to create digital object repositories, which they can then use, and reuse, in their teaching. Bob is showing an example of pictures relating to Rome created by a fine arts students. An interesting aspect here is that the IDETS team provide a ‘rationale’ to the library to start the creation of such a collection, and the integration into learning objects.

This ‘Rome’ project is 461 images in DigiTool (in different versions), built a UI in Flash, which interacts with DigiTool with web-servercies API, php and the DigiTool PID, with supplementary material from Grove dictionary of art – a prototype seems to be available at Some impressive stuff, here, and our recent online picture gallery could have been done using exactly this technique – if we had DigiTool. Bob, said that they could probably develop several such resources per year, but in comparisson to the number of faculty and courses, this is just a drop in the ocean…

My talk seemed to go down well, and I’ll try to blog to detail later.

The other panelists (including me) are now covering work at their instiutions (so excuse any sudden cut off as I have to speak!)

Firstly, Peter van Boheemen (Wageningen University, NL). They use Blackboard and Sharepoint. They have developed a function to allow the building of ‘course libraries’ in the library system. When the tutor links from the course to the library system, they are automatically logged in as the owning ‘course’ rather than a personal login. This sounds a bit like the work done at Westminster a few years ago, where MetaLib e-shelves were setup for each course…

Secondly, Kingsley Boulton (York, UK). They are using Sentient ‘Discover’ product for reading lists, and Blackboard for a VLE (apparently the Sentient product is now being acquired by another company, and will be called just ‘Discover’). They are only adding reading lists for those courses that have a Blackboard presence. They have a ‘e-learning librarian’ to do this work. They went live in January this year, and used a Blackboard ‘building block’ (developed by Sentient) to integrate. The reading lists are completely controlled by the academics. The idea is that library staff will rely very much on the lists in Discover to get details of what academics need purchasing, moving to short loan etc. User feedback at York has been generally positive. York are also looking at a building block to integrate MetaLib with Blackboard using a ‘Building block’ and the x-server. The aim is to have single sign-on between Blackboard and MetaLib, and be able to present specific MetaLib resources based on the student enrolments.

Thirdly, Bob Gerrity (Boston College, USA) is going to talk about using DigiTool with the University environment. At Boston College there is a IDETS (Instructional Design/e-Teaching Services) who work along side library staff to deliver services for e-learning. Bob’s got an interesting list of strengths for the library team, and IDETS team – something they try to keep in mind as they work together.

Boston College have already done hooks to library resources (from WebCT) – search forms, deep linking, links to subject specialists – this sounds very similar to what we have done at RHUL so far.

However, they are now starting to use DigiTool to provide a platform for faculty members to create digital object repositories, which they can then use, and reuse, in their teaching. Bob is showing an example of pictures relating to Rome created by a fine arts students. An interesting aspect here is that the IDETS team provide a ‘rationale’ to the library to start the creation of such a collection, and the integration into learning objects.

This ‘Rome’ project is 461 images in DigiTool (in different versions), built a UI in Flash, which interacts with DigiTool with web-servercies API, php and the DigiTool PID, with supplementary material from Grove dictionary of art – a prototype seems to be available at Some impressive stuff, here, and our recent online picture gallery could have been done using exactly this technique – if we had DigiTool. Bob, said that they could probably develop several such resources per year, but in comparisson to the number of faculty and courses, this is just a drop in the ocean…

My talk seemed to go down well, and I’ll try to blog to detail later.

This morning the first presentation is by Jon Dunn from Indiana University, and as you might guess is about working with the Sakai course management system.

Later in the session, I’m contributing to the panel session, describing and discussing what we have done, and aim to do with Moodle and library systems at RHUL.

Jon is starting with an overview of course management systems or VLEs. He is making a strong case that it is essential for the library to go out and work with the VLEs – we cannot rely on the fact that students, or even lecturers, will come to the library (perhaps especially ‘virtually’), and so we must invest in integration between the VLEs and the library systems and resources.

At Indiana, they have been working specifically with Sakai (I hadn’t realised that Sakai comes from Hiroyuki Sakai – from the Iron Chef TV programme!)

Anyway Sakai began in 2003, with 4 institutions which wanted more control over institutional CMS and felt it would be more cost effective to develop in house, than to pay a vendor. From this Sakai has grown into a group of 105 paying partners, although only 8 institutions are in production, with 11 more scheduled for Q3 2006.

Sakaibrary is (unsuprisingly) a project to integrate library systems/resources with Sakai. It is a partnership between Indiana University and University of Michigan, and runs until 2008.

The project goals are:

To build tools to provide seamless integration of content from licensed library datases within Sakai for instructors
Leverage existing library technology infrastructure
Prototype functionality for librarians to present content in Sakai and students to discover licensed content within Sakai
Engage librarians, students and faculty in this process

Once the tools have been developed they will be released as open source (although you’ll obviously need Sakai to use them)

The project started with some tasks that different stake holders might do – a selection of these are:

Lecturers

Search for articles
Import references from Refworks etc
Create persistent links to articles for student access
Create customized search box or ‘canned search’ for students (e.g. automatically add keywords to search to keep it on topic)
Obtain assistance from library

Students

Access article found by instructor
Search for articles
Export references
Obtain help from librarian

Librarians

Create links to appropriate databases
Create customized search boxes
Create help guides and suggestions

Sakai defines ‘tools’ (frontend interface) and ‘services’ (backend functionality). Sakaibrary is working on the folloing deliverables:

Phase 1:
Library search tool – based on existing work (Twin Peaks Navigator, OKI), but to leverage metasearch technology.
Citation management tool

Phase 2:
Subject research guides

Jon is just demonstrating the integration. A nice interface add library resource appears as an option within the html wysiwyg editor, and this pops up a search window (just prototype at the moment), and then inserts a selected link into the course material. It wasn’t entirely clear what html was behind the link, but sounds like they are using a combination of OpenURL and RDF.

For the citation management tool, they have only got prototype screens at the moment, but the basic idea is to be able to import citations, or search for citations using library resources, and then add to your citation list.

The subject research guides will provide functionality like ‘constrained searches’ where searches are limited to specific areas to help guide the students to appropriate resources.

The tehcnologies being used at Indiana are Unicorn, MuseGlobal, and SFX. At Michigan they use Aleph, MetaLib and SFX. This means the project is determined to get the functionality across different library systems. To acheive this they are using OSID (Open Service Interface Definition) which came out of the MIT OKI.

The key OSID function is the respository query and delivery (OKI Repository OSID). Since this was already implemented in Sakai, it was a logical choice. University of Michigan is using the x-server to access MetaLib functions, using CQL as internal query representation, and OKI Repository OSID for communication with Sakai. Similar work happening with MuseGlobal at Indiana.

Some interesting comments from Michigan regarding the MetaLib X-Server – some XML encoding and validation issues, and the lack of ‘appending merge’ option – to try to allow paging through merged results, even though you don’t retrieve all results at once.

With SFX, they have found it fine, but apparently have found it isn’t possible to customise the menus based on the fact the Source is Sakai – but I’m really suprised, as this should be possible?

The Search and citation management should be ready for testing Q3 2006, and the subject research guide tool in Spring 2007, with Open source release in mid to late 2007.

Some unanswered questions are:
Who will use search within the CMS? Possibility that lecturers won’t, as they have references from other sources already, which is one reason to support import from other areas.
Will metasearching ever be good enough?
Will OpenURL link resolution work consistently enough for full text access?
How will the Sakai development model work in libraries?
Will other institutions step up to tackle other library integration?

Some really good stuff here – and perhaps raising the bar on what we need to look at in terms of integrating citation level information into the VLE – so far we have only discussed integrating service/resource level information (e.g. ScienceDirect, Web of Science, Nature Online) – because this seems easier.

It would also be very interesting to look at how much (if at all) Moodle supports OKI Repository OSID, and whether we can use this for our own integrations. I really liked the way that adding a library resource to the html, was integrated into the html editor – very nice, and perhaps extendable? (it wasn’t clear how this was acheived)

This morning the first presentation is by Jon Dunn from Indiana University, and as you might guess is about working with the Sakai course management system.

Later in the session, I’m contributing to the panel session, describing and discussing what we have done, and aim to do with Moodle and library systems at RHUL.

Jon is starting with an overview of course management systems or VLEs. He is making a strong case that it is essential for the library to go out and work with the VLEs – we cannot rely on the fact that students, or even lecturers, will come to the library (perhaps especially ‘virtually’), and so we must invest in integration between the VLEs and the library systems and resources.

At Indiana, they have been working specifically with Sakai (I hadn’t realised that Sakai comes from Hiroyuki Sakai – from the Iron Chef TV programme!)

Anyway Sakai began in 2003, with 4 institutions which wanted more control over institutional CMS and felt it would be more cost effective to develop in house, than to pay a vendor. From this Sakai has grown into a group of 105 paying partners, although only 8 institutions are in production, with 11 more scheduled for Q3 2006.

Sakaibrary is (unsuprisingly) a project to integrate library systems/resources with Sakai. It is a partnership between Indiana University and University of Michigan, and runs until 2008.

The project goals are:

To build tools to provide seamless integration of content from licensed library datases within Sakai for instructors
Leverage existing library technology infrastructure
Prototype functionality for librarians to present content in Sakai and students to discover licensed content within Sakai
Engage librarians, students and faculty in this process

Once the tools have been developed they will be released as open source (although you’ll obviously need Sakai to use them)

The project started with some tasks that different stake holders might do – a selection of these are:

Lecturers

Search for articles
Import references from Refworks etc
Create persistent links to articles for student access
Create customized search box or ‘canned search’ for students (e.g. automatically add keywords to search to keep it on topic)
Obtain assistance from library

Students

Access article found by instructor
Search for articles
Export references
Obtain help from librarian

Librarians

Create links to appropriate databases
Create customized search boxes
Create help guides and suggestions

Sakai defines ‘tools’ (frontend interface) and ‘services’ (backend functionality). Sakaibrary is working on the folloing deliverables:

Phase 1:
Library search tool – based on existing work (Twin Peaks Navigator, OKI), but to leverage metasearch technology.
Citation management tool

Phase 2:
Subject research guides

Jon is just demonstrating the integration. A nice interface add library resource appears as an option within the html wysiwyg editor, and this pops up a search window (just prototype at the moment), and then inserts a selected link into the course material. It wasn’t entirely clear what html was behind the link, but sounds like they are using a combination of OpenURL and RDF.

For the citation management tool, they have only got prototype screens at the moment, but the basic idea is to be able to import citations, or search for citations using library resources, and then add to your citation list.

The subject research guides will provide functionality like ‘constrained searches’ where searches are limited to specific areas to help guide the students to appropriate resources.

The tehcnologies being used at Indiana are Unicorn, MuseGlobal, and SFX. At Michigan they use Aleph, MetaLib and SFX. This means the project is determined to get the functionality across different library systems. To acheive this they are using OSID (Open Service Interface Definition) which came out of the MIT OKI.

The key OSID function is the respository query and delivery (OKI Repository OSID). Since this was already implemented in Sakai, it was a logical choice. University of Michigan is using the x-server to access MetaLib functions, using CQL as internal query representation, and OKI Repository OSID for communication with Sakai. Similar work happening with MuseGlobal at Indiana.

Some interesting comments from Michigan regarding the MetaLib X-Server – some XML encoding and validation issues, and the lack of ‘appending merge’ option – to try to allow paging through merged results, even though you don’t retrieve all results at once.

With SFX, they have found it fine, but apparently have found it isn’t possible to customise the menus based on the fact the Source is Sakai – but I’m really suprised, as this should be possible?

The Search and citation management should be ready for testing Q3 2006, and the subject research guide tool in Spring 2007, with Open source release in mid to late 2007.

Some unanswered questions are:
Who will use search within the CMS? Possibility that lecturers won’t, as they have references from other sources already, which is one reason to support import from other areas.
Will metasearching ever be good enough?
Will OpenURL link resolution work consistently enough for full text access?
How will the Sakai development model work in libraries?
Will other institutions step up to tackle other library integration?

Some really good stuff here – and perhaps raising the bar on what we need to look at in terms of integrating citation level information into the VLE – so far we have only discussed integrating service/resource level information (e.g. ScienceDirect, Web of Science, Nature Online) – because this seems easier.

It would also be very interesting to look at how much (if at all) Moodle supports OKI Repository OSID, and whether we can use this for our own integrations. I really liked the way that adding a library resource to the html, was integrated into the html editor – very nice, and perhaps extendable? (it wasn’t clear how this was acheived)

Because of the bimonthly updates to SFX, the ‘versioning’ seems to matter a lot less than with other products. Perhaps this is also because the functionality is simple, and stable.

Ex Libris are planning to move to releasing semi-monthly (rather than the current monthly) KB updates (although you can take them when you are ready – you don’t have to take them every fortnight if you don’t want!)

Some other recent and ongoing enhancements are:
A-Z title list improvements
OpenURL 1.0 SAP2 XML
Better linking for conference proceedings (starting with IEEE)
Enhancements to KB data

SFX is now moving beyond just being a ‘link server’, this is being supported by work on:

Support for ANSI/NISO Z39.88-200 (OpenURL 1.0) – SAP2 including multiple Context Objects
Dynamic A-Z list
Citation Linker
KBTools: exports, collection development analysis
Integration with ERM
Strong consortium support
Support for ONIX SOH 1.0 (Onix Serial Online Holdings) – for initial and ongoing subscription maitenance

This last one is something I need to investigate more – but presumably this will allow our suppliers to supply machine readable holdings information which SFX can consume, and use to update our local thresholds – this is potentially incredibly powerful, but I’m a bit wary about the accuracy of supplier data, as in the past it hasn’t always been that good (even though you’d really expect them to know)

The figures for the KB are impressive with almost half a million object portfolios over about 1500 targets. We are also seeing growing numbers of e-book targets – 56 targets, and 360,000 objects.

A real challenge here is how our users access e-books (see my earlier post relating to the possible need for a special interest group for this). If we loaded extensive e-book records into our catalogue, they would swamp our existing catalogue (e.g. EEBO roughly 100,000 records, our catalogue roughly 300,000 records).

I think that probably we would expect our users to access services like EEBO direct, or via MetaLib. However, as Nick pointed out in a comment to a previous post, there is also a move towards buying individual titles – so more like a ‘print’ replacement – in this case, where else are they going to live except the library catalogue?

SFX 4

SFX 4 is on its way (currently slated for Q3 2007), and some highlights include:

Enhancements to the admin interface including task-based wizards which are intuitive and workflow oriented
Further integration with Verde and sharing of KB data
Better integration of print holdings (definitely very interested in this
Improvement of end-user ‘menu’ linking experience
Further international support
Improved management of statistics
Towards KBUpdate ‘on-demand’

Some disappointment here that the integration of print holdings isn’t already here. We saw a presentation (just mentioned again in fact) last year where a library had done it, but by going deep into the SFX code and changing it – which Ex Libris won’t support, and I don’t want to do.

Last year I blogged about this, and it’s nice to note that 62 libraries have taken up this free service from Ex Libris.

MetaLib v4 is due out later this year, so some highlights here…

Accessibility

This has long been a bone of contention with MetaLib (and other Ex Libris products), and finally we are seeing the fruits of long conversations with Ex Libris about this. With ML 4 we have coming:
Section 508 Compliance
JAWS, Dragon and wide range of browsers tested
Use of HTML, CSS Validity (HTMS 4.01 Transitional, CSS 2.0)
Simplicity – ‘lighten’ HTML (reduced source code by 60-70%!), and reduce use of JS
Ease of Customisation – to reduce TCO

At the same time Ex Libris have endeavoured to keep the user interface exactly the same as it is in v3.x. There are some slight changes, but these are mainly in the areas of new options (e.g. non-js version). Some changes that may be noticed by the end user are that the pop-ups are now different, and can be resized, moved etc. (unlike currently), also when you leave MetaLib, then a message informs them that this is happening with a customisable message.

Enhanced Search results

This is a major change in v4. There will be the ability to cluster and ‘facet’ search results. This should enable users to analyze results at a glance, and to expose new or related topics to the user – to enable navigation and further refining of results.

‘Clusters’ are derived from text mining of a user’s search results records. Ex Libris are using licensed clustering technology from Vivissimo. Note that Clusters are not taken from a taxonomy, and are built ‘on the fly’. The clusters are presented as an expandable/collapsible hierarchy.

Facets on the otherhand are ‘pre-selected’ (by Ex Libris), and are built on structured fields from the retrieved records – for v4, Ex L are providing Author, Publication date and Journal title facets.

From the users point of view, ‘clusters’ and ‘facets’ are presented in the same way, so they aren’t aware of any difference and don’t have to worry about the semantic differentiation above! The user just sees the option to view records by topic, date, author or journal title.

KnowledgeBase Enhancements

Over the last year between 250 and 300 have been added to the central KB over the last year – with around a third of these coming from user driven requests.

Ex Libris prefer to work with resources that support z39.50 or XML standard gateways – as this helps maintain a low TCO.

Rollout of ML4

Quite a lot of enhancement originally intended for ML4 have actually been released in service packs to v3.13. This includes:

X-Server enhancements (sign up for some details of user implementations in the ‘Innovative Users of Metalib’ Webinar on the 18th October).
Statistics and Reports (I haven’t noticed any improvements here – perhaps we need to look at these again?)
MetaIndex (which in our case, we have not got)

ML4 will start by release to Ex Libris offices and distributors in December 2006, and will start to be available to ‘early adopters’ in early 2007. There will be an ‘upgrade express’ kit, which will allow customer’s to apply the upgrade without impacting on the ‘live’ version (given enough disk space presumably).

The future

Even with the upcoming developments with Primo, MetaLib is stil a key component in Ex Libris’ overall strategy. The areas of focus for development are:

MetaLib at ‘point of need’ – user focused and integrated into other environments such as VLEs, Portals and of course Primo
Lowering TCO for MetaLib

Oren Beit-Arie is the Chief Strategy Officer at Ex Libris, and is going to talk about the company’s vision and strategy, and also will talk briefly about a new product – Primo.

Originally a ‘Library’ was a place. Library management was about managing the physical space, and the physical stock – providing access to the physical collection was the challenge.

Ex Libris started in 1980 with the Aleph library system, which was focussed on the challenges of managing physical collections.

However, in recent years, there has been an exponential growth of electronic resources, and now libraries are managing both a physical collection, and electronic resources. Associated with this is a shift from owning to licensing resources for libraries.

Comparing ARL stats of total materials expenditure, and e-resources expenditure, the 11 year compound annual growth rate up to 2004 has been 6.3% for overall expenditure and 30.2% for e-resources. In 2004 e-resources expenditure represted over 30% of total expenditure.

So – this has led to some new challenges/needs:

Linking between web of resources (the appropriate copy problem)
Simultaneous search across disparate collections and systems
Management – new models of acquisition and licensing

And of course (this being a supplier presentation!) Ex Libris had developed systems to help with these problems – SFX, MetaLib and Verde respectively – which have been introduced from 2000 onwards.

Cynicism aside, SFX certainly led the way in terms of OpenURL resolvers. It is also worth noting that all three products are the result of close collaboration with the Ex Libris user community – a strong recommendation for both the company, and the work of the user groups and user community.

There is also an increase in digital assets being created – perhaps this applies especially to the University market, where digitisation of readings, exam papers, pre- and post-prints, archives, and increasingly ‘born digital’ material – all of the above, and learning objects, web pages etc. etc.

Oren proposes that the ‘big picture’ can be split into things ‘under library control’ (probably, but not necessarily, ‘local’), and those ‘not under library control’ – e.g. e-books, e-journals, databases – often licensed resources. Oren also notes that this isn’t really the big picture – this is the ‘library’ picture – if we look at the user, they will be drawing a huge range of information sources and services.

Also, we are now in an environment where many users are also creators (although recent figures indicate that this isn’t a huge percentage)

All this leads Ex Libris to the following strategy:

1 – the end-user: provide a system and services that will enable libraries to expose content and services to users where and when they need it and in the way they expect it

Library content and services should be offered to users in context – so promote interoperability, open interfaces.

I’m really hoping that we are in the market for a Digital Object Repository in the next 12-18 months. We are currently implementing Moodle, and I’m keen to see this underpinned by a good repository. Of course, being an Ex Libris customer, DigiTool is one possibility on my mind.

In this session, it is interesting to see some uses of DigiTool in the real world (links should go to the DigiTool instance where possible):

E-Theses, E-Prints (Liverpool John Moores) – this includes what looks like a nice work flow and user interface for E-prints
Online exam papers (University of Melbourne)
Online photos (Coldharbour)
Cultural Heritage collections (quite a lot of places – Spertus College, Boston College, Boston Public Library, AIATSIS, BVB)

The Cultural Herigate collections have a huge amount of variety – books (METS), pictures (tiff, jpeg), audio and video

EAD – not to manage the archive, but to store and present the information from the archive.

Catalog enrichment (HBZ) – they are using it to scan and store the Tables of Contents for more than 300,000 books and then use these to enrich the metadata available via the catalogue. (looks like this is in test at the moment – http://193.30.112.126:8881)

Preservation (HBZ again) – using it to store web pages in an archive (also looks like it is in test – http://193.30.112.125:8881 – search for “loreley”)

Overall this has been a good overview of real life uses of DigiTool – it’s good to see a real body of use developing. I have a particular interest in how DigiTool might integrate with VLEs or Course Management systems – for example, to work as a repository for SCORM or IMS CP objects.

In terms of future developments, Ex Libris are currently working on an improved OAI interface, synchronisation between Aleph and DigiTool, Versioning, improved METS viewer, URN and persistent identifiers, additional metadata support – e.g. VRA and MODS.

Coming out of a discussion over coffee, a question – are the search functions within VLEs good enough? My recent experience with Moodle is that as soon as you have any substantial amount of content, the search isn’t good enough (try searching at http://moodle.org)

So – if we were going to index parts of our VLE, where should we start? Discussions, content…?