Liveblog notes from an IET Technology Coffee Morning on 21 July 2010. Hassan Sheikh and Owen Smith from the OU Library present.
Why does the Library need to have systems management? Because it has a lot of systems: some legacy, some newer.
The users are key, especially the students – they use the systems to access e-resources. But also course teams, ALs who do that and come in to the physical library.
Key systems they’re working on at the moment:
- vertical search
- Library Management System
- Library website (migration to Drupal from Knowledge Network, Web 2.0, mobiles, web services/feeds, personalisation)
- integration with other systems (finance, DRM)
- ORO (ePrints)
- Linked data and reference management (TELSTAR and LUCERO projects)
- authentication (SAMS, Shibboleth, EZ Proxy)
- eBook readers.
Current federated search uses connectors to many online services, but they’re generated dynamically, so it’s slow, and if the connector is broken you don’t get access. Also it’s a separate system to all the other search systems and interfaces – leads to a lot of confusion from students. Google generation expectation is a single system. So a new vertical search system – locally index the metadata, so faster, more coherent, and easier to provide relevancy ranking. Using EBSCO Discovery Service, should be implemented in the Autumn.
OU Library website – 400k page visits, 60k visitors per month. Most popular: home page, search, databases, eJournals and eBooks. And more than 91k unique logins to eResources every month. Developing new Mobile version – three devices model – MIT/NCSU mobile web: low-end (basic web-enabled phones); middle end (smart phones, BlackBerry), high end (touch screen, Android/iPhone).
More work in progress on the website. Drupal version of the site. Search terms are not very clever – top search (15%) is for ‘one-stop search’, which is the default text in the box. Next most popular is 10% searching for a course code, which won’t work for searching external journals. They clearly don’t understand.
Resources by course code lookup – show the user resources related to the course they’re interested in. Courses can span more than one library subject; user might not know where to look.
OPAC/Voyager modules – show user what books they’ve borrowed, requested, or had recalled. Clever stuff behind the scenes to link OUCU credentials to Library barcode.
Vertical search modules.
Question: Amazon. Looking for camping lantern, scrolled down and found ‘people who searched for this ended up buying this’. Anything to help me if I get the search term wrong?
Yes, will show you ‘people actually read’. Uses metadata scoring based on what people read for more than 30s.
Question: Often happens when UK English phrase is different from the American one.
Can do spelling correction and UK vs US dictionary.
Scott: Is this done as a widget so you can embed it?
Is done as RSS.
Scott: Would be good in e.g. Moodle environment
Are working on integrating it like that.
Scott: Like the tag cloud of related searches from other people on the course, becomes very powerful.
Then asked audience: what would be cool?
Scott: Do you have library lookup working with Voyager catalogue?
No – only people who use Voyager are on campus.
Scott: I keep my things I want to read in Amazon. Have to take the reference out of that, put it in to the catalogue. Voyager can query that instance on your Amazon page (your wishlist) shows you what’s in that Library.
Keren: Can do this and specify your local library. Allow students to say I have access to e.g. University of Bedford library and Bedford public library, and search that.
Scott: Workaround is to use Worldcat. That is a challenge because the OPACs on the back end are all different.
Liam: Providing an API so tinkerers can experiment with library data?
Liam: Like scan a barcode to see if the book’s there. Query the catalogue and get a JSON object back.
Scott: That data ‘students on this course queried with these terms’, that’s gold.
Hope to release all this data, also who’s been searching for my book, who’s visited the library and done the work. Are my researchers doing stuff in the library doing things. Lots of work done when Huddersfield released their circulation data. But ours isn’t interesting because we don’t have much footfall.
Scott: Course, subject to Dewey mapping, would be interesting.
Yes. With TELSTAR, want to integrate that in to our search – that has a list of books/articles to read. Do the vertical search to tell you where things are. Give course code data to the world.
Pauline: How far out are we getting to end users? Have been researching how students and academics engage with resources. Coming out – we have wonderful systems, which can reduce the user’s work to almost nothing. The users tend to have their own systems, styles and patterns of accessing the resources. They have their own seemingly-working styles. Experts would know there are limitations – e.g. no quality, or critical rating. So they could have better outputs. How can you try to accommodate them, get their input?
We try to ask them, but our userbase, it’s not just students but everyone – three or four different levels of students. Student wants ebook to read for their course. Academic wants articles to write course. Researcher wants everything. We have to deal with publishers, they hate telling us anything about what we’ve given to people. They’ll say 20,000 articles read this month, but don’t get a list of what they were, to tie down to who that was, without a lot of hacking about with the authentication. Ideally we’d know who they were, what they’ve searched for, what they’ve read. So could say, read all of these, which ones were useful.
Scott: Vast majority of stuff they have to authenticated to see? (Yes) So you can create a different front end, and do the tracking yourself rather than relying on goodwill of publishers.
Hassan: The logs (EZProxy etc) are all in very different formats, have to do a lot of processing to understand what’s happening.
Hassan: Integration with DRM, would be good to show what they can do with things – restriction on printing, reading, downloading. Working on that but is far down the line.
Pauline: Using the authentication to track sounds interesting. I was searching journals from home, had to login. I could see the abstract but not get full text. Thought OU Library doesn’t subscribe. Then came here, thought to get through British Library, but just checked, and there it was. So do we have to levels of authentication? We cut out so many people who don’t get in to the system? A Francis & Taylor journal – Information and something or other.
Could’ve been a problem with the journal yesterday. Or a rights issue. On campus can see everything we subscribe to, some things are limited – ALs can only see a subset. Staff can see everything, students is a separate subset. External content has very complex licences – can show to X people of certain types, at certain times of the day, within certain time periods.
Anna: Can you report it to the helpdesk?
Keren: Maybe vertical search will report that.
Vertical search/federated search only shows you what you can see/subscribed to. But publisher’s search shows it all.
Scott: Might help to have option – show me what I can get at vs show me everything.
Voyager shows results for full collection, some of it you can just take out, some of it you have to get permission to access – e.g. multimedia resource. As an end-user do you need to see that?
Someone: Ablity to limit and understanding that is important.
Click here for more makes it easier.
Someone else: Am a regional academic based in Bristol. Is it possible that I’ll get different results if I search in my office in Bristol I’ll get different results if I search here?
Yes. If you use 360 search. But vertical search will give you same results every time.
Also depends on time – overnight the index might update. But same time on same day.
Anna: Example of journal was probably a problem. Shouldn’t happen you don’t get access to things in some places when you are allowed in others.
Regional academic: A lot of staff also have a student identity, may see a different search if you log in with the student identity?
The aim is to have resources that students need always available. Things they may not need, or isn’t suitable, is available to staff members. Want to be consistent, give them what they need straight away. But possible to do discovery/learn more about this afterwards.
Q: Students have their resources on Student Home for the course. If they go to the Library website, do you show them the resources before they search?
Give option – if search for course code, show those resources. People sometimes put in course names to library catalogues to see whether they want to study a course.
Also My References – course team creates list of resources with the library, go in to a page of resources on the course website. Can use that as a starting point to search to do other things.
This work by Doug Clow is copyright but licenced under a Creative Commons BY Licence.
No further permission needed to reuse or remix (with attribution), but it’s nice to be notified if you do use it.