Archive for the ‘Dave Pattern’ tag
The case for opening up library data: #jiscmosaic at Wolves
I made the (deceptively long) journey over to the city campus of the University of Wolverhampton yesterday (18 Nov 2009) for the concluding JISC MOSAIC project event. I was without an Internet connection all day so wasn’t able to waste time contribute to the backchannel by tweeting from the event, so here’s my writeup.
Interlude 1: what’s all this about?
MOSAIC stands for ‘making our shared activity information count‘ (but took its name from an earlier project with the acronym TILE. Lots of TILEs = a MOSAIC. Geddit?). It’s being funded by JISC, and it’s “investigating the technical feasibility, service value and issues around exploiting activity data”. For ‘activity data’, read (in the main) library book-circulation data. Say the JISC:
“MOSAIC aims to build on [the TILE project] by aggregating library activity data from several institutions and making it available for re-use and experimentation. The Talis podcast with Dave [Pattern, University of Huddersfield] provides further background.”
I.e., if lots of university libraries shared their anonymised circulation data in a common format, what Web2.0-type-o’-magic could we build on top of that data, and how would it benefit our users?
We’ve been invited to contribute some of Lincoln’s own (anonymised) Horizon circulation data to this project. I’ll write more about our own involvement in a future blog post.
After introductions, we were welcomed to Wolverhampton by Fiona Parsons (Director of Learning Centres at Wolves, and vice-chair of SCONUL), then talked through the project’s progress-to-date by David Kay of Sero Consulting Ltd, the lead institution in the project. There was discussion about some of the challenges that have faced institutions wanting to contribute their own activity data – in particular, the difficulties involved in extracting the data from different models of LMS, and institutional concerns about privacy, utility, cost, and the ownership and re-use of ‘their’ data*.
The keynote presentation came from Paul Miller of Cloud of Data Ltd - ‘Activity data and the global information economy: the who, what, when, where, how, why of an emerging future’. I hope Paul (or the MOSAIC project team) will put his slides online soon – it’s well worth a read.
In the meantime you might like to look at Paul’s blog: http://cloudofdata.com/
Next came coffee and the first of two breakout sessions. In a group with Ken Chad, Jill Griffiths (MMU), Alex Parker (So’ton University) and a guy from Wolverhampton**. The breakout session was titled ‘Being Practical’ and we were tasked to come up with real-life use-case scenarios for one or more different types of HE library user.
We focused on the undergraduate, and spent a bit of time discussing students’ trust in the reading materials given to them by their institution, students’ reading behaviour and information literacy, how improvements to library processes (including considerations of VfM) impacted on the student experience, how to ’sell’ the utility of library usage data to universities, and particularly students’ motivation to read in particular ways.
This led us to what was meant to be our use-case scenario, which we were a bit nervous about, so we rather tentatively posed it as a question instead! (I’ve had to rephrase it from memory below because I left our original notes in Wolverhampton, but this was the gist…)
Interlude 2: our use-case scenario: “Read Your Way to a First”
Can we use library activity data to learn anything about the reading behaviour of students who get higher degree classifications that we could use to inform the reading behaviour of all students?
Obviously, there are some huge questions and potential dangers hidden behind that innocuous question, and a hypothesis (i.e., that there’s some relationship between your reading behaviour and the degree you end up with) that would have to be tested first. It sparked some lively discussion in the run up to lunch, as did the other groups’ use-case scenarios for undergrads, researchers, academic staff, library directors, and developers of web applications for libraries.
Over lunch (good sandwiches; always important) most people sat down to watch a pre-recorded slideshow presentation with voiceover from an absent Dave Pattern about how Huddersfield are really using real circulation data to really improve their students’ experiences of the library. (It’s an adaptation of a similar presentation I saw the flesh-and-blood D.P. deliver at last year’s Mash Oop North event.)
I recommend you take a look at his presentation - it’s only 17 minutes long and it’ll be time well spent…
After lunch we ran through a series of ’perspectives on the problem space’ – some excellent and genuinely thought-provoking presentations from Mark van Harmelen, Ken Chad, Paul Walk, Jenny Craven & Jill Griffiths of MMU, and Helen Harrop of Sero, which led to some equally thought-provoking discussions.
Once the presentations are available online, I’ll post a link.
At about 3:15pm the group broke up for tea and started another breakout session & discussions - unfortunately I had to take this opportunity to go and tackle the M6.
Interlude 3: what next for us?
As I said, I’ll blog more about how Lincoln can and will [fingers crossed] contribute some of our own anonymised Horizon circ. data (also, I hope, some e-resource usage data) to this project before 2009 is out. In the meantime, the data’s still out there, just waiting for someone to come along and build innovative library services on top of it…
Ten useful links:
- About the event at Wolverhampton
- MOSAIC – JISC project page
- MOSAIC web pages (SERO)
- Free book usage data from the University of Huddersfield
- Talis podcast with Dave Pattern on sharing usage data
- MOSAIC data wiki and an API
- The MOSAIC developer competition, the results and Alex Parker’s winning entry
- Paul Miller’s Cloud of Data blog
- Dave Pattern’s presentation
- #jiscmosaic thread on Twitter
* I’d argue that it’s not libraries’ data at all. It’s our users‘ data; we’re just keeping it safe for them.
We’re not un-hot
Dave Pattern’s HotStuff 2.0 blog analyses more than 800 ‘biblioblogs’ in an attempt to discover new and/or interesting topics…
Today, the L&LR staff blog is…
loading…
Sharing book usage data for the benefit of all
The University of Huddersfield have made their circulation usage and recommendation data free and reusable for the benefit of other libraries. Library systems manager, Dave Pattern, envisions a “book recommendation service that makes Amazon’s look amateurish”, made possible through more and more libraries doing the same.
Dave asks:
“whether or not you can augment this with data from your own library. If you can’t, I want to know what the barriers to sharing are. Then I want to know how we can break down those barriers.”
Mashed Library ‘08 – proper writeup
I promised to turn my as-it-happened notes from last week’s Mashed Library event into a proper, readable account of the day. A few other attendees have already written up their respective ‘takes’, including Tony Hirst, Paul Walk, Jo Alcock and organiser Owen Stephens, to whom credit is due.
Me, looking wide-eyed at Mashed Library '08. Photo by Dave Pattern.
Here’s my effort…
The day started interestingly enough, spotting Red Kites from the window of the train on the way down to London. Once I’d located Birkbeck College and the room under the library which was to be the venue for the day, it was straight on to connecting to Birkbeck’s e wireless network (the first challenge!).
This was more of an “unconference” than a traditional, speaker-driven event, with everyone tapping away at their laptops and nipping across the room to compare notes (and to get more coffee – the day’s catering was excellent!) throughout.
The first demo was from Rob Styles from Talis, who gave a quick explanation of some of the possibilities of their Talis Platform for searching and “augmenting” library (and other) data – a Harry Potter-themed example (Hogwarts’ finest kept cropping up through the day, which perhaps gives you an insight into the reading habits of systems-librarian types?) of a search of Talis’ test bib. data store created an RSS feed which was then ‘dipped’ (‘augmented’) with book cover images from another data store.
I certainly think we should take a look at Talis Platform and consider joining to their development network. I’ve already signed up to their mailing list / Google Group.
Some useful links:
Next, Tony Hirst from the OU ran through a number of freely available mashup-inspiring tools that can inspire people who “aren’t happy on the command line” (i.e. normal library workers!) to play with library data. Tony knocked together a series of lightning mashups, retrieving and presenting book cover images and Amazon reviews for a given ISBN, using a number of sites including:
- Amazon Web Services Made Simple
- Google Docs – spreadsheets can be used to scrape HTML pages and manipulate the data using familiar cell/formula operations
- Yahoo Pipes and the similar pipes.deri.org
- ThingISBN
I’ve already started mucking around with Yahoo Pipes to repurpose the data that’s locked up within services such as the A-to-Z; to create bespoke combined blog feeds for my own use, and (tentatively) on the XML output of HiP. I don’t doubt there’s more we could do (particularly involving the readily-available book-cover image collections that are out there), but as with a lot of this 2.0 stuff, it’s where to start?
Timm-Martin Siewert from Ex Libris was next up, demo-ing EL Commons, their rough equivalent to the Talis Platform – with the important difference that, at the minute, access is restricted to Ex Libris customers only. EL Commons provides usable APIs for various EL services, including Primo, Metalib, Aleph, etc. Timm was closely followed by Mark Allcock of OCLC, who showed a number of that organisation’s developer tools, including the WorldCat Developers’ Network.
I was impressed (and slightly overwhelmed!) by the range and number of OCLC’s reusable services which Mark mentioned, including registries of copyright information, journal ISSN history, and more. Some are restricted to OCLC members (although the Developers’ Network is not), but others are open to all:
- WorldCat Affiliate Services (not restricted to OCLC members)
Verdict – merits further investigation.
If your head is beginning to swim at this point, you’ve about reached the stage that I was at by now!
Aaaaaaaagh: code!!!!! Photo by Dave Pattern.
Last up before lunch was Ashley Sanders, a programmer at MIMAS. He spoke briefly and tantalisingly about some significant developments on the horizon for both COPAC and ZETOC, including better persistence of searches (presumably so they could be used in RSS feeds etc.) and a “modify record” tool that may be opened up to the community on the live service[s], once the question of authentication can be sorted.
Gratefully, on my part, we broke for lunch* (which was very good, again).
There were a number of interesting chats going on. In particular, around the OCLC’s “powergrab” (not my words!) of reuse-rights relating to catalogue records which it provides (a debate I’ve been following with mild interest) vs. a more “free”, creative-commons-Web2.0-hippy approach to making bibliographic data available for sharing. It was getting quite heated at one point (in a polite, librarianly way, of course).
I got more coffee.
The next bit of the day was the bit I struggled with most. Perhaps I’ve been conditioned to crave structure(!), but as people milled around, discussing ideas and working in small groups, I started to feel a bit on the periphery. As it was, I collared Rob Talis and got him to explain (patiently… and patiently over again…) how Talis Platform might be used in practice. After I’d asked a few daft questions it started to become a bit clearer (Rob – if you’re reading this, that’s not any reflection on your explanation, only on the time it takes for new ideas to sink in to my brain). At the time, I wrote:
“hadn’t realised this was entirely free (as developer) and unconnected to Talis LMS. I’m suffering from library-terminology disjoint syndrome at this point – different language / terms of reference.”
As with all of this stuff, I’m convinced you can only start to expose the value by getting stuck in and trying things out. Library2.0-in-theory doesn’t work.
My sanity was restored considerably at this point by Dave Pattern, who came over for a chat, and by the cakes and scones that started to appear (with more coffee). Dave and I talked about OPAC-wrapper / social-discovery-layer and metasearch applications (much, much more of which later…), the gathering and intelligent use of statistics, and Dave’s valuable tweaks and improvements to HiP.
My tags as I blogged on the day. Photo by Dave Pattern.
Credit also due to Dave for his photos, some of which I’ve used here.
The day wound up with one last presentation: this time from Paul Bevan, of the National Library of Wales. His was a talk I could identify with! Themes relating to geographic isolation (of Aberystwyth); a shift to virtual rather than physical visits to the library; institutional ‘buy-in’ and acceptance of the use of free 3rd-party social media applications; accessibility and the differing needs of different groups of users.
Finally, Owen got everyone in the room to shout out what they’d been working on through the day. I admitted that I was nervous about the idea that I was meant to have achieved something by the end of the day (ha, ha) and that I didn’t really have anything to show. Others had been working on some really interesting ideas, including lots of use of Yahoo Pipes to bring together information from Amazon, Wikipedia, Talis Platform et al., and a couple looking at deriving geodata from MARC geographic name headings, to display catalogue records (e.g. for items relating to a particular town or city) via a map interface.
Incidentally, I had my first real ‘Mashed’ idea on the train on the way home. I’m working on something quite interesting, and if it looks as if it’s got legs, I’ll post it here.
This was a fantastic day in many ways, with 1,000 ideas a second being generated. I did find it slightly frustrating… I just need time and a quiet room to get my brain around this kind of thing, and I felt that 80% was passing me by. The challenge, I think, for L&LR, will be to identify which services and techniques have real potential for sharpening up our own service, and putting appropriate time aside to develop them properly. Otherwise, I fear, it’ll ever be seen as a slightly eccentric hobby (“You do what with pipes!?”)…
Last set of links:
- Mashed Library ‘08 – Ning (social networking) site
- Google Alerts RSS feed
- Facebook site I set up after the event
I’ll certainly be looking forward to Mashed Library ‘09, and if possible I think two or three of us should go next time.
Paul
—–
*No, I don’t know why I call it “lunch” at a conference, when it’s “dinner” in all other situations.
Change to e-book links on HiP
Chris L. and I have made a couple of changes to the way e-book links work on the public catalogue.
I’ve put a message on the public blog about the most obvious difference – MyiLibrary titles now open up in a new browser window – this seemed to be something that annoyed a lot of users (at least – a few Holbeach students complained about it…) because they couldn’t get back to the catalogue using the ‘back’ arrow.
Horizon now also handles the Athens authentication from MARC 856 $u in a slightly different way. We’ve made that change so that we can get the MARC records for the new ebrary e-books onto Horizon with the minimum of fuss, but it shouldn’t affect what users see when they log in to MyiLibrary (or ebrary) via HiP.
(You might spot that, since the change, links ro existing MyiLibrary titles in 856 are getting a double-dose of Athens authentication… this doesn’t seem to cause a problem [cue emails from Eduserv telling me why it's a terrible idea!] and Chris, Di, Bev, Jill and I will have a little project after Christmas to tidy up the existing MyiLibrary MARC records, to bring them in line with the newer ebrary records, and to put a stop to the duplicated Athens authentication within the links.)
I have to give credit here to Dave Pattern @ Huddersfield, who pointed the way forward to do this within Horizon(!) when we were chatting in London a week or so ago. Thanks Dave; works a treat!
Google Book Search Data API
Google have launched an API for their Book Search, and Dave Pattern is already wondering how to nail it to his OPAC…
Paul
Huddersfield’s EcOPACs
Dave Pattern, systems librarian at Huddersfield (whose blog I follow religiously) has just posted about how he’s potentially saving his university thousands of pounds a year and at the same time keeping Huddersfield green by replacing their ’beige boxes’ with low-energy-use OPACs.
Paul (taking a break from late-night MSc work!)
