Things 14 and 15 of 23: LibraryThing or LibraryUserThing?

The most recent task for the Cambridge 23 Thingers has been to look at LibraryThing. I’ve been vaguely aware of this site for some time, and it does include one of the neatest tricks for data mining usage patterns I’ve seen, the UnSuggester. This works best if you enter fairly well-known novels, and goes beyond simple high-culture low-culture divides, to tell you that if, like me, you like Dorothy L Sayer’s ‘Gaudy Night’ you should not just skip the work of Chuck Palahniuk (which may be obvious), but also Paulo Coelho and Jodi Picoult (which is less obvious). But it was only 23 Things that has actually prompted me to get stuck in and try other aspects of LibraryThing.

So my personal library, or at least a very small part of it, is now up there. I decided to give LibraryThing a little test by including a range of my academic books, including some with no ISBNs and one foreign text (Regine Le Jan’s Famille et Pouvoir). Overall, it didn’t do badly. I found everything on there already (including Regine’s book), and though most of my books were fairly unusual, a surprising number of people have Kennedy’s Revised Latin Primer. The biggest problem is finding multiple variants when you haven’t got an ISBN, such as with my 1955 Penguin Classics version of Bede. The fact that it’s possible to edit details of any of the books you’ve found is handy, but also sends shivers up my cataloguer’s spine (because it affects records globally).

What happens when you do have your (personal) collection of books on LibraryThing? From a quick look, the site is a lot shorter on reviews than I have expected. For example, there are no user reviews for Rosamond McKitterick’s Charlemagne: the formation of a European identity, while there are (completely different) ones on both Amazon US and Amazon UK.

LibraryThing’s tags are not much use, but the recommendations aren’t at all bad. Take, for example J W C Wand, A History of the early church to AD 500. The tags are “a4, christianity, church history, churches, early church, early church history, ecclesiology, history, roman empire”, which tell you little that isn’t already in the title. But the recommendations include the Penguin translation of Eusebius and another volume of translated sources from the early church. (I’m not convinced, however, that the recommendations are better than Amazon’s ‘Customers who bought this item also bought’). It’s also noticeable that even with only 13 books added, the ‘members with your books’ feature has already highlighted as similar both Another Damned Medievalist and Curt Emanuel, both of whom I already knew from their
medievalist
blogs.

There are also some nifty book-centred tools on the site. The links from each title to Google Books is potentially very handy. The site is also strong on search links to both booksellers and libraries (slightly ironically, it may thus be most useful for books that aren’t actually yours).

So would I find it useful personally? Not for my own historical work. I need reference management software (I currently use EndNote) which can deal with journal articles as well as books, which is under my complete bibliographical control, and which allows me to format the output in multiple ways. Even the Google Books link feature (which does look useful) could potentially be replicated via the ability to set up personalised libraries on Google Books.

For about 90% of the time, meanwhile, I don’t need to have records of the other books I own – I can remember them or find them on the shelves. The one time it would come in useful would be in a bookshop. If I had a mobile phone I could surf the web on (I currently don’t), I could check exactly which works of Margary Allingham or Terry Pratchett I had before purchasing another (and so could any other relatives who want to give me books).

In contrast, LibraryThing does make a lot of sense for many amateurs (and I mean that term in a positive sense, for people who love books and reading). It’s far simpler to use than setting up your own database, and it also provides suggestions and the chance to connect with like-minded people. If you’re a public librarian working in reader development I can see it being an important tool. And it would also be handy for small libraries that can’t afford library software. (I once did a library catalogue for a small firm of solicitors in Excel, because that was what they had).

For larger libraries, I’m a lot less convinced that it’s useful. I had a look at the catalogue of the San Francisco State University Library which includes tag clouds generated from LibraryThing. Here are a couple of screen shots pulled from a search on Chris Wickham’s books:

Thing 14 no 1

The tag cloud for ‘Early medieval Italy’ had 13 terms, and the vast majority of them are either superfluous (‘history’, early medieval’, ‘Italy’) or too general to be useful (‘power’, ‘sociology’). Only two (‘barbarians’ and ‘late antiquity’) actually add useful information.

Thing 14 no 2

‘The mountains and the city: the Tuscan Apennines in the early Middle Ages’ has one tag: ‘early middle ages’.

If you’re reserving a place on the catalogue page for tags and thus adding to both processing time and onscreen ‘clutter’, you need to be getting something useful from this. I’m really not convinced that for academic non-fiction you do.

Some libraries have also put their new books up on LibraryThing, for example, Nuffield College, Oxford. My question is whether the extra work involved is worth it? Given that Nuffield already have a new books list on their own site, who is their target user on LibraryThing? Who wants to know about the new books they get, but is not sufficiently aware of them as a library already to visit their site? I presume that much of the process of adding new acquisitions to LibraryThing can be automated, with a string of ISBNs passed to the site, but I suspect not every new item in many academic libraries would have an ISBN, and it would still be quite a lot of work setting up the feed initially. For some kinds of library and library user, I can see the big attraction of LibraryThing. For academics and academic libraries, I’m much less convinced so far.

Advertisements

6 thoughts on “Things 14 and 15 of 23: LibraryThing or LibraryUserThing?

  1. I’ve enjoyed LT myself though some of the discussions can get a bit tedious. You’re absolutely correct though – I consider it “social cataloging.” I can’t see how it would be highly useful for an academic. It should come as no shock that I use it substantially for my medieval hobby, not at all for my profession.

    One of the features it has had that is regrettably down right now is something called “tag watch.” Basically, when someone else enters a book with one of the tags you’ve selected, you get a notification. I can ignore 90% of that but every now and then something useful comes up.

    And the real fun is “raiding” someone else’s collection. I particularly enjoyed diving into Edwards James’. Unfortunately that leads to a huge wish list I’ll never fill.

    Like

  2. An interesting post. I hope you don’t mind but I’ve sent a link to this to Nuffield College Library (I used to work there) and asked whether anyone there wanted to comment. Perhaps they’ll give an insight on why they decided to have two acquisition feeds.

    Like

  3. Thanks to Kirsty Taylor for letting me know about this – I will respond on behalf of Nuffield College Library, where I am Librarian. We set up our accessions on LibraryThing mainly as a ‘getaround’ – a way of being able to get RSS feeds into college members’ inboxes, since our college website is awaiting a full overhaul/update at present. We have had good feedback from it. We are trying a variety of ways of getting our messages across – we also notify our updated accessions lists on Twitter, which in turn is linked to our Facebook site. We felt that we were working hard to put our accessions lists on our website every week and no-one was looking at them (library website? Yawn….), so for very little extra work, we could try a variety of approaches to cater for all preferences. So, our target users are still college members, and the fact that it is on a ‘public’ site a bit incidental.

    Like

  4. As a fellow lover of Gaudy Night (well, of all Dorothy Sayers, really), I could have told you to avoid Jodi Picoult, but I don’t really see the connections between the two that the UnSuggester is picking up on (though I never actually finished the Picoult book I started). I’ve never really thought of LibraryThing as a research/cataloguing tool – more a site where I can find fellow readers’ recommendations along the lines of ‘if you liked this, you may also like…’ One thing I like is the I See Dead People’s Books project (Legacy Libraries), which is cataloguing the libraries of historical figures like Margaret of York.

    Like

  5. I have nothing to add except that today’s “reCAPTCHA” for me to comment consists of the phrase “arbitrary Philosophical,” and it would be a shame to leave that truism (tautology?) unused.

    Oh yes – count me among the Sayersphiles (and a former subscriber to the “Dorothy-L” list).

    Like

  6. Thanks for the comments – I’ve just done another post on social citation sharing that ends up discussing LibraryThing again (and thanks to Curt for mentioning the tag watch option, which is handy to know).

    Kirsty and Elizabeth – it’s interesting to hear the background to the Nuffield decision. It reminds me how often library decisions that may look odd to an outsider actually make sense given particular constraints on them, and their past histories. Libraries don’t often have a blank sheet to play with when developing their services.

    Bavardess – the Unsuggester (and the Suggester) doesn’t work by knowing anything about the books concerned, just about the people who own/read them. As I understand it, it starts from knowing that X% of LibraryThing users have a particular title Y. From that it can calculate the expected number of LibraryThingers with Gaudy Night in their library who also have Title Y. It then compares this expected number with the actual number of people who hold both titles and picks the ones with the largest discrepancy downwards – e.g. you’d expect 41 people who read Gaudy Night to have title Y in their collection, but there are actually none. This works surprisingly well, because I suspect many people, even across a number of genres, are quite consistent in their preferences on the style of writing, the type of characters, the underlying world view etc.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s