L&LR staff blog

Spotted on the Eduserv data messenger:

Emerald is adding a further 12 new titles to the portfolio in 2010. The new titles will be available as part of Emerald Management Plus – subscribers will automatically gain access to new and acquired titles as and when they are added.

To celebrate the changes Emerald is making, they are giving away an HP laptop which is wireless enabled and has Vista pre-installed! For your chance to win the laptop, simply complete the following caption:

Emerald Management eJournals are…

Email your caption with your name, job title, institution name and country to quiz@emeraldinsight.com.

The most original caption will win. The competition ends on 30 September.

Hopefully, you should have noticed a few recent changes to the Electronic Journals A-to-Z.

• First, we’ve all but completed the summer campaign to improve the accuracy of the 60-odd individual e-journal package holding files. Di, Adele, Carole, Phil, Elif and I have been working through each package in turn and using the most up-to-date information from the publisher/provider to improve the A-to-Z holdings. One or two inconsistencies still to iron out, and it’ll never be entirely bug-free (please carry on sending any mistakes that you spot to acquisitions@lincoln.ac.uk), but the quality of the data for our managed packages should be significantly better than it’s ever been. This review process is going to be [it's going to have to be] an annual event.
• Second, the ‘New e-journals at Lincoln‘ RSS feed has been updated through the summer to reflect the changes. Noting that EBSCO’s terms and conditions for use of the A-to-Z state that we ”may not distribute, encourage or allow distribution of [...] data updates“, I’m going to be restricting the RSS feed to just those packages where we, not EBSCO, are the source of the updated information (i.e. where we are managing the title lists directly). This includes the packages providing access to our individual title-level subscriptions: i.e. the ‘important stuff’.
• Third, and last of all, as I discussed with several people in my workshops at the L&LR awayday in June, I’ve revamped the A-to-Z home page ready for the start of the 09/10 session…
  As discussed in those workshops, I’ve taken a ‘just do it!’ approach. It’s not the finished article by any means – I’m relying on your initial and ongoing criticism, comments and ideas to take it forward – but we agreed: better to make changes then tweak as appropriate, rather than discuss ouselves into inaction.

(Image: the new A-to-Z homepage)

Compare with the old home page (below):

(Image: the old A-to-Z homepage)

The changes / new features include (hopefully these will ring bells with those of you who took part in the discussions):

• Generally – the aim has been to make the site more clean and simple.

1. The page ‘header’ – extraneous information (and slightly excitable formatting!) removed – just the name of the application (there was some discussion about dropping ’Electronic Journals A-to-Z’ in favour of ’E-journals Catalogue’ or similar, but now’s probably not the time for that) and the strapline with the number of titles.
2. The navigation tabs – I had hoped (and people had agreed) to reduce these to a minimum, replacing most of them (advanced search, subjects, etc.) with lower-profile links further down the page, but I encountered a problem – if I use A-to-Z admin to ‘hide’ a tab, it becomes impossible to link to that tab elsewhere. This is weird, and unfortunate in the extreme – unfortunately it’s something we’re stuck with for the time being. Inevitably, I’ve ended up adding to the number of tabs – as well as a new tab for the new home page, there’s also now a ‘Log in‘ tab which routes the user through Athens (and back to the homepage), which ought to help us to troubleshoot users’ full-text access problems.
3. A big, unmissable search box. Needs a better search button graphic, but it does at least now include a search tip (to subtly reinforce the idea that users should be searching for journal titles here).
4. See also the sentence on using the e-Library for initial subject keyword searching  – could we / should we expand on this?
5. More options – links to the other A-to-Z pages. I had intended these to replace the tabs entirely, but see (2). Maybe, given that, we don’t actually need these extra links?
6. Still to come – a screencast video showing the user how to use the A-to-Z in the context of a literature search. I’ll blog when that’s ready. Also need to revamp the ‘Help’ tab more generally.

So – what do people think?

I’ve finally got a year’s worth of e-book usage data from both the ebrary and MyiLibrary platforms.

I’ve been commenting on the growth in usage throughout the year (since Michelle formally launched ebrary Academic Complete back in January), so here (without comment) are all the stats from August 2008 and July 2009.

Graph: e-book usage as a % of total usage (print+e).

[...]

[...]

[...]

…alright, so I can’t resist commenting.

It looks as if we are definitely levelling out at the 25% mark – i.e., one e-book session for every three paper loans. It’ll be interesting to see what happens in September – will this 3:1 level of preference carry on into the new academic year?

My assertion that this translates to “one electronic loan for every three print” has been rightly challenged – after all, a new e-book session will be recorded every time a user ‘opens’ an e-book, whereas the same can’t be said for print loans, where the ‘loan’ is only registered once, irrespective of whether the reader opens the book thirty times, once, or not at all. Really, we need a formula along the lines of:

S = rL

Where S = a single user session, L = a print loan, and r = the number of times on average a student reads a library book they’ve borrowed. Anyone care to give me a figure for r?

What’s particularly interesting to me is that this growth has happened with – in the grand scheme of things – little in the way of overt publicity. I know that we’ve all pointed readers to e-books where they’ve been available, and I’m sure the Academic Subject Librarians have made sure their respective subjects were aware, but I’ve a hunch that the real driver has been the existence of e-book MARC records on Horizon, allowing readers to discover the e-books serendipitously. What would happen if we did the same for our 44,000 e-journals?

Back in November, I blogged about the disproportionate nature of our e-journal usage, once you break it down by title (if you remember, one single title is responsible for 5% of all of our e-journal hits, while a small European country’s worth of journals are never accessed at all).

While looking at the A-to-Z usage stats some more, I started to notice some other unusual features – not, this time, relating to individual titles, but to the usage of entire packages of journals. (“Packages” for A-to-Z purposes, are things like Academic Search Elite, HeinOnline, SwetsWise, etc.)

Here’s a loverly pie chart:

Starting in the top-right hand corner (coloured light blue), 16% of our e-journal usage can be ascribed to titles from (the EBSCOhost package) Academic Search Elite. No surprise there.

Going clockwise, the next slice (the dark-pink/magenta/purply-brown/mauve/maroon??? one. I don’t do colours! This website calls it “rouge”…) represents the usage of our SwetsWise subscriptions – 13%.

Then, the first surprise: the first ‘exploded’ slice in the bottom right of the chart (cream) represents the usage of the various free or open-access e-journal packages which are listed on the A-to-Z.

12% of all our usage – one eighth – derives from journals which we don’t pay for. Most of this is from journals listed in the EBSCO Open Access Journals package.

Continuing clockwise round the three bottom slices (pale blue, purple, sort-of-peachy), we encounter Business Source Premier (10% of total usage), ScienceDirect (8%), and ABI/INFORM Global (7%).

The next exploded slice (blue) was another surprise to me… this 7% of usage is for the data-collection on the A-to-Z of links to the OPAC for our current print journal subscriptions.

…I’ll let that sink in… one in every fifteen hits on the “Electronic” Journals A-to-Z is actually someone looking for a print journal.

I only really added the current subs data to the A-to-Z as an intellectual exercise, and it’s overdue a review (they are created manually, and of course they go out of date quickly). Perhaps I should be taking it more seriously.

Going up the left-hand edge of the chart are:

• IngentaConnect (4%)
• SPORTDiscus with Full Text (4%)
• Sage Publications (2%)
• Factiva (2%)
• Oxford University Press (2%)
• HeinOnline Law Journals Library (2%)

The Rub’ al-Khali at the top of the chart (11%) is everything else. That’s everything else including Emerald, including JSTOR, and including LexisNexis. Who knew?

…whether the subject is food or finance, or – in theory – any keyword you want, you can filter the RSS feed of new e-journals by constructing a link that looks like this:

• http://pipes.yahoo.com/pipes/pipe.run?_id=mDjFxR1R3hGHC2SH0z6skA&Subject=XXXXXXXXX

Where “XXXXXXXXX” is your subject word.

That link creates html which you could link to within Blackboard (see illustration below – click for bigger). Or, to get the titles as RSS, just add &_render=rss to the end of the URL.

This is done using Yahoo! Pipes.

Many libraries, in HE and elsewhere, now display book cover images on their OPACs. Example (sorry, Dave): webcat.hud.ac.uk ~ no doubt, this is something we should be doing in the near future.

~~~

In the meantime, I was thinking how desirable it would be to similarly enhance the Electronic Journals A-to-Z with images of the journals it contains. But, while databases of book-cover image files are relatively well-known and accessible [Amazon, LibraryThing], I’ve not been able to find any such general collection of journal covers available for re-use by libraries – is it that periodicals are more jealously guarded by individual publishers?

Or is it that they’re out there somewhere, but that I’ve not been able to find them?

What I have discovered is that there’s at least one provider of academic e-journals that makes its cover images freely available for re-use: Elsevier, purveyors of ScienceDirect, whose cover image files “may be used in systems in which Elsevier Science material is offered to end users“.

I’ve used Elsevier’s directory of journal cover images (http://www.elsevier.com/wps/find/products.cws_home/journal_cover_images_intro) to create a “Note” within A-to-Z admin, assigned to all ScienceDirect titles, which uses the ‘Link URL‘ fuction to display a journal cover image, matching the ISSN of the journal, on the live A-to-Z site.

Each image links to its journal home page.

~~~

Problem #1 – if there is no image matching a particular ISSN, the A-to-Z still tries to display an image, and the result is a “red X” missing-image icon (). Can we live with that, for a relatively small number of journals, at least until Elsevier add the missing images to their site (hint, hint)? I’ve tried to filter out the missing images using Yahoo! Pipes, but haven’t managed to get it to work, yet.

Problem #2 – more of a question, really: can I abstract this method to pull in images from other publishers’ sites? And which publishers actively allow this, or provide a dedicated service as Elsevier do?

You can now subscribe to my RSS feed of new e-journal titles via email, should you so wish.

• Here’s a link to the email sign-up form.

I prefer to keep up-to-date with RSS feeds using Google Reader (I don’t want or need anything more in my Outlook inbox!), but if email’s your bag and/or you prefer to have all your information in one place, then at least there’s now the option to add these to your inbox.

I have a feeling that there’s just too much in this feed (c.90 new titles added over the Bank Holiday weekend) to be useful as a public i.e. library-user-focused service. Too many titles, no subject filtering (yet…), too much (for want of a better word) noise in the form of low-academic-interest titles added to packages such as Factiva.

But I do think this could be a really useful current-awareness tool for Academic Subject Librarians.

In February, I wrote about  how I’d added more than 5,000 ticTOCs RSS feeds to the A-to-Z, so that researchers can easily find and subscribe to Table-of-Contents (ToC) updates from journals in their field. I think this is an indispensable current awareness service, and I’m pleased it’s being promoted at the ‘Working Smarter With the Web‘ workshops.

(An aside: that ticTOCs data is overdue to be re-generated and reloaded onto the A-to-Z. This is next on the long list of e-journal packages for the “e team” to review.)

That’s all well and good, but there’s a clear need for a current awareness service ‘one level above’ the individual ToCs. That is, we could (and should) be providing an RSS feed of new e-journal titles.

At the moment, the EBSCO A-to-Z platform doesn’t provide this feature (I really do hope it’s something that EBSCO are considering developing themselves), and so in the past we’ve had to resort to emails, spreadsheets, etc.—all with varying degrees of success—just to keep people up to date with what new e-journals have become available.

~~~

So, I’ve created a rough-and-ready working demo of a ‘New e-journals at Lincoln‘ RSS feed for the A-to-Z. It’s not supported by EBSCO, so it’ll require a bit of manual intervention to keep it running (it’ll be a nice five-minute daily task for the “e team” to take on!), but has the potential to be a really important enhancement to the e-journals service.

Next time you go to the A-to-Z site, you’ll see a new tab – next to the existing ‘Titles’ tab – which I’ve labelled ‘New Titles (RSS)‘. This new tab links to a page containing information about the RSS feed, with a prominent link to the feed itself, plus a display of the 10 most-recently-added titles in the feed.

The links in the ‘ten most recent…’ list should go directly to the A-to-Z record for that title. The feed link itself you can paste into your favourite feed reader (I use Google Reader). I’m working on enabling email subscription to the feed. We should also be able to take this feed and filter/manipulate it to create (for example) subject-specific lists in Blackboard.

I’ve also made the feed autodiscoverable (c.f. my earlier post), so it should display in the browser toolbar (this will vary between browsers).

This is how the feed is auto-discovered by Firefox browser.

I still need to do some work on creating better documentation for users of the feed, and for the process of adding new titles within L&LR.

I’ll also need to do a bit of testing to make sure everything’s working as it should be, and iron out my feed-validation bugs.

When the time is right, I’ll launch the service more publicly.

Paul

~~~

P.S. Here’s how I did it:

1. The first stage is the most clunky – it involves getting information about the newly-added titles out of the closed box that is the Electronic Journals A-to-Z. EBSCO provide a daily report (available only to administrators) of “Titles added to Packages in my Collection“. This data can be exported on a daily basis in tab-delimited format.
   
2. Next, I created a spreadsheet using Google Docs and published it to the Web. The tabbed data from the A-to-Z admin report is pasted in to the spreadsheet, so it becomes openly accessible. Using Google Docs also means that I can invite other people to act as editors of the spreadsheet, and create public web forms to add titles that don’t appear in EBSCO’s daily report (not all of our e-journal packages are managed by EBSCO, so the data will need supplementing when we add custom titles).
3. The data output of the Google spreadsheet is then fed into Yahoo! Pipes, an excellent application for ‘mashing up’ and processing data. You can inspect the Yahoo! pipe for yourself, at: http://pipes.yahoo.com/lincoln/newejournals. If you create an account on Yahoo!, you can clone this (or any) pipe and modify it to create your own data mashup.
   
4. The Yahoo! pipe takes the raw Google spreadsheet data, filters it according to various rules, and exports a valid * RSS feed of new items, in reverse order by date.
5. * Well, it’s nearly valid. I need to do a bit of cleaning up of the pipe (I did say this was a working demo), mainly so I can pass the feed through Feedburner, which will allow email subscriptions. But it works well enough!
6. Finally, I created the new tab on the A-to-Z public site, and added the details to the page (the ‘ten most recent’ list was created using feed2js). The code for the autodiscoverable feed was pasted into the top of our customised A-to-Z template.

~~~

I’ve just added the title data for the Periodicals Archive Online (ProQuest by way of JISC Collections) to the A-to-Z, and it’s tipped our number of unique titles (subscription and open-access) to 37,001.

A minor Friday milestone.

Caveats/notes:

1. Authentication of the P.A.O. is by IP address / Shibboleth, but I haven’t fully tested it for the usual off-campus problems.
2. I haven’t yet added the P.A.O. to the e-Library pages as a separate database (and I won’t until I’ve done #1).
3. It could dip back below 37,000 as titles are automatically removed from managed collections.
4. It doesn’t matter anyway.

More than 36,000 e-journal titles are listed on our A-to-Z, but student and staff interest is not spread equally between those titles.

Plotting a graph of the individual usage of each title shows a sharp ’spike’ of concentrated use within a very small number of titles, and a very long tail of thousands of titles which are barely used, if at all. In fact, as a raw graph it’s almost impossible to read, as all the action takes place at the extreme right-hand edge (click the image for bigger):

Some comparisons to put the distribution of usage into context:

A single title—the British Journal of Social Work—is responsible for 5% of all of our e-journal hits via the A-to-Z (5,185 hits in the last year). This is fairly remarkable in itself. What’s the secret there?

Anyway, imagine that this one e-journal (0.0027% of the total number of titles) is represented by the area of the West Common in Lincoln: about 100 hectares. Remember that 5% of all our A-to-Z usage is here.

Photo by Julian Beckton

The next 10% of A-to-Z usage can be accounted for by just 14 titles. If one title is the Lincoln West Common, then 14 cover an area about the size of the London Borough of Islington (or Windermere, if you prefer): approx. 6 square miles. These 14 titles each receive between 500 and 5,000 hits/year.

(Scale outline maps reproduced from Ordnance Survey map data by permission of the Ordnance Survey © Crown copyright 2001.)

The next 366 titles get fewer than 500 hits/year (39% of total usage) They tot up to just a bit bigger than Rutland, the smallest county in England (not counting the made-up counties), at 147 square miles.

We’re getting in to the beginning of the long tail now: 2,365 titles (the East Riding, including Hull, pushing 1,000 square miles), with 50 hits/year or fewer - this represents the next 34% of usage.

And 6,122 titles; 5 hits/year or fewer; south of the river to Lincolnshire, at 2,687 square miles the second-largest English county. Just 12% of our total A-to-Z usage from 17% of our titles.

Finally, 75% of all our e-journals - 27,745 titles in total – are never used via the A-to-Z.

That’s zero A-to-Z hits in the last year, although it’s quite possible they’ve been accessed via the native database or publisher interface, or through tools such as Google Scholar. To represent those 27,745 unused titles at the same scale, we’d have to use…

…Belgium, which won’t fit on this page.

Imagine that: a medium-sized European country full of e-journals, and none of them used. Quel dommage! Or, if you prefer, Een welk medelijden*.

*Reckons Babelfish. I don’t speak Flemish.