Today I attended the UC&R (University, College & Research Group) Wales event ‘The changing face of library catalogues: new OPAC developments’. It went beyond OPACs though to encompass the wider world of resource discovery systems, and it was refreshing that all the talks were from the point of view of the HE libraries implementing each system. That gave a useful perspective on the who, what, when and why – with warts where necessary – that you would not get just from talks by commercial suppliers. An impressive thing about the programme was that there were going to be seven talks, and each was about a different answer to the question “What tools should we provide to help our users find the resources they need?” Although the event was about HE implementations it was good to catch up with colleagues from HE and FE, the National Library and academic departments. I have typed up my notes below, and in the interests of being extra helpful have tried to link to each resource at its first mention.
11:00-11:30 Primo at Aberystwyth University / Christiane Kloos
In my Aberystwyth University role I have been involved with this system so was familiar with how we got to where we are now, but for everyone else it was useful that Christiane began with the history. Times are changing and libraries are looking at keeping up with user expectations. Our Voyager library system was not going to be upgraded by Ex Libris in the future so we needed something else if we wanted to upgrade. Aberystwyth had SFX for e-journals and Metalib as a federated search tool – all Ex Libris products. Although we had been an early adopter of Shibboleth, Voyager had its own authentication system. The multiplicity of authentication systems and different access points was not ideal – it made more sense to have a single, intuitive user interface. One point for resource discovery, and only one user interface to keep up to date.
The current Primo implementation
Christiane’s team had looked at a number of options: Aquabrowser (they felt it was too public library-focused); Summon (had a nice interface); Vufind (an open source system but it wouldn’t seem to work with our Voyager system); WorldCat Local (promising but too US-focused); and Primo. At the end of the day they all look and function in similar ways. The advantage for us of using Primo was that it was from the same vendor as our other systems, which should help with interoperability.
The project started in May 2010 with a small team, aiming for an August 2010 launch. Shibboleth authentication was incorporated first; then making an OPAC available via Primo; and most recently work on implementing Primo Central. There were delays so the system was launched as a beta for the start of term 2010, and work is still ongoing. The main problems were with some OPAC functions, full-text availability, and statistics. On the 2nd March 2011 the team made the default search a combined Primo Central and local holdings.
There has been lots of feedback (some surveys offered free printer credit as an incentive). Currently users can select whether they want to use Voyager or Primo, and about twice as many users seem to be preferring Voyager, but Primo usage is increasing. Christiane gave a live demo to show the way that Primo integrates the patron functions. You can view the current Primo implementation here.
11:30-12:00 Encore at Bangor University / Mieko Yamaguchi
At Bangor they wanted to make the OPAC more interesting by adding features such as tagging, reviews, personalisation etc. Encore was the product they chose. However this was where one of the recurring themes of the day came up – naming. Mieko began by pointing out that Encore is the commercial name for the resource discovery product, but the name would never be mentioned to users. There are pros and cons to both branding options, for example using the commercial name may mean that people who have used the system elsewhere could be happy to realise they have access to a system they are familiar with. On the other hand, the chance to brand something locally as either a new product or as an update to an existing portal can be beneficial for user continuity and reassurance.
Naming issues came up again when Mieko pointed out the difficulties in explaining to users the difference between an Encore ‘quick search’ (the Encore search box can be embedded anywhere) and ‘simple keyword search’ or OPAC search (which would be a search of the OPAC via their LMS Millennium). Both allow a keyword search and both can find books in the library, yet they search different systems. Mieko pointed out that many other libraries that have Encore offer both search options, one via Encore and one via a classic catalogue link. This can be confusing for users. At Bangor they wanted Encore to be the default, but this is a work in process, since Encore is not fully customisable, and issues include not being able to use it to search for a classmark (a common request, as anyone who works in a library will be aware!)
Mieko showed screens of their implementation, with a tag cloud on the right and refining facets on the left. An innovation was that within records users can click on the location for a link to a pinned Google map (identified by coordinates rather than postcode for greater accuracy); their system also links to Google reviews. They will soon add Encore Synergy for full text linking, and include harvested records from the repository. Mieko also showed screens of other institutions’ implementations before discussing user and staff feedback. In the future Mieko wants to better integrate Encore and the OPAC.
12:00-12:15 Questions and discussion
- There was some discussion of what mechanisms were used for feedback. In many cases it was a link on the first page of the catalogue.
- The issue of naming came up again, trying to explain to users what they are searching. For example Ex Libris do not allow purchasers of Primo Central to list the databases that a user is searching when they use it – so how can a searcher know whether they have included all the major databases in their literature review? We always have to consider two audiences. We have one group that wants a quick and easy way of finding materials; and also a smaller audience of ‘expert’ searchers who need advanced options and have to know what they have searched so they can see when they have exhausted all the main possibilities. Hiding the search targets from them is incredibly frustrating.
- The question was also asked: to what extent are users involved in the purchase and development of new systems? Librarians were accused of being ‘notorious’ for buying systems and setting things up then presenting it as a finished thing, where it is too late for users to affect things. This is often due to tight timescales and budgets that have to be spent by a certain date, but where possible involving learners can be valuable.
1:30-2:00 Implementing SUMMON at UWIC / Julie Neenan
Julie began by taking us through the early stages of UWIC’s resource discovery journey, and it was a story that is familiar to many librarians. In 2009 they had multiple interfaces and routes to content (Athens, Google, OPAC, repository etc) which led to a confusing user discovery experience and access routes, especially off-campus. Julie applied Zipf’s Law to the experience, the interpretation being that users will take the route of least effort to achieve anything – another familiar story! UWIC discovered that their LMS Talis Prism 2 would cease to be developed by Talis and they would have to pay a lot for a minor upgrade to Prism 3 if they wanted to stick to the same system.
There were also lots of reports coming out at the time: the Google Generation report, HE e-books observatory report, various RIN reports (‘Mind the skills gap‘, ‘Overcoming barriers‘ and ‘E-journals use, value and impact‘), the JISC Service Provider Interface Study and so on. These raised many questions about how our users find materials.
UWIC staff decided that for 2009/10 they would just concentrate on getting the basics right. They moved resource information for students from the external website (which became a marketing tool rather than useful resource – the third familiar story…) into an intranet portal, called Electronic Library. For 2010/11 they went on to look at authentication and moved from Athens to Shibboleth and EZproxy. Julie said it was one of the best things they did – there was a fair bit of work involved at the start but it was well worth it, and established good foundations to build on.
UWIC already had some Serials Solutions products (360 Link resolver, 360 Core, 360 MARC Updates). In August 2010 they added 360 Resource Manager, set up a harvesting mechanism and started extracting data out of Talis Prism so that the new Summon system they had selected could be a one-stop-shop, with search boxes that can be embedded anywhere. Summon seemed to be ideal – the IT department was busy with large projects (familiar story #4) but with Summon the library could mostly do the work themselves, bringing print and e-resources together and allowing search and retrieval in one interface with faceted options. They hoped to maximise usage of their resources in this way.
The project has been held back a bit by problems extracting data out of the Talis Prism system. There was a large cost if they wanted Talis to do it, but the system is not set up to make it easy for you to do it yourself. Therefore they are going ahead with the other steps first, and will add OPAC data later. They might just launch Summon as an e-resource portal in September 2011 until they can incorporate the OPAC functions with the extracted Talis data and maximise the value of their MARC data.
Pitfalls they have faced include the Talis data extraction problems, and some problems of interconnection between Proquest and EBSCO which inhibits resource discoverability. Commercial competition and tight reins on metadata harvesting was a theme discussed a few times during the day. Looking to the future Julie was interested in the use of statistics comparing educational attainment to usage of e-resources and the library: there is obviously huge potential there for libraries.
2:00-2:30 Vufind at Swansea University / Paul Johnson
This talk had to be cancelled which was a shame since I am interested in open source software and how far you can get using it. Particularly when there was so much talk during the day of closed systems, lack of interoperability, commercial rivalry, companies refusing metadata harvesting, inability to export your own data, high costs paid to companies, lack of support when it has been paid for etc.
2:30-3:00 WorldCat Local for Trinity Saint David / Alison Harding
Many of Trinity Saint David’s decisions had been driven by the merger of two HEIs. As a result they had two LMS: Voyager and Horizon (which they still have because they have not implemented WorldCat Local yet – the beta is to launch in April 2011). They wanted a new solution, since both LMS were at the end of their life and were comparatively clunky. Trinity Saint David looked at Aquabrowser, Vufind and WorldCat Local as possible replacements. WorldCat Local allows users to search multiple catalogues simultaneously plus other systems and some e-resources and it has a Web 2.0-type interface. The other systems offer this too but in the end they chose WorldCat Local because they favoured the support options.
Trinity Saint David currently uses OpenAthens but would love to move to Shibboleth since Athens administration takes up so much time. They will still use the separate LMS systems behind the scenes, but will offer one OPAC search for the users (even if they have to resort to native LMS interfaces for patron functions). So far they have had issues with items that have no ISBNs (e.g. the oldest books pre-1840s; artefact collections), and the quality of the bibliographic records varies – something that becomes more visible with advanced resource discovery interfaces. WorldCat Local is not currently used heavily in the UK, so most of the live examples are US-only.
3:00-3:30 Questions and discussion
- A lengthy discussion of the issues of commercial metadata hoarding took place. For example resources like Westlaw and LexisNexis don’t allow metadata harvesting, which makes their products less flexible and harder to integrate into a single streamlined process for users. Likewise Ex Libris won’t allow libraries to announce the things being searched in Primo Central, as discussed earlier. The librarians all agreed that these kind of approaches are behind the times.
- There was some discussion of open source software. The point was made that whether you go with commercial or open source, you can pay for support (or not). For example, PTFS Europe support the implementation of open-source LMS such as Koha and Evergreen. Therefore the old argument that open source software has no support is no longer true.
- We discussed the sharing of cataloguing records. I pointed out that every institution is currently paying for them (whether as an imported purchase or salaries for local cataloguing); also many institutions pay for expensive conversion projects to bring records up to date. This is in many ways a duplication of cost. This could be a clear win for sharing of records in a simpler way, whether just within HEIs (organised by WHELF) or across all libraries (perhaps supported by CyMAL). It makes sense to share; at least for records that are not gained under restrictive licence agreements.
Overall it was an interesting and thought-provoking day, many thanks to UC&R Wales. The fact that every talk covered a different system meant that as well as serving as an overview of some of the major choices, the day illustrated common themes and messages, most of which I have mentioned somewhere above. One issue that we did not discuss much was user training. Some people claim that modern resource discovery interfaces should be self explanatory ‘like Google’. However any power user knows that even Google is not straightforward, and to understand things such as Creative Commons searches requires extra guidance. I am a fan of screencasts as a way of showing the main features. <controversy> They tend to be quicker to work through than long text pages with screenshots, and more attractive too. Why implement a snazzy new system then parade it around with only old-school library culture? </controversy>
I’ll link to the presentations once UC&R Wales makes them available.
A Beginners’ Guide to Screencasting: A Lunch Time Byte delivered to your desk
