1. ISBN 2. Web Service 3. ??? 4. Profit

April 17, 2007 1:50 PM

I need to know which book-related web service/API is best (most complete archive, most up-to-date, most metadata) for looking up information about books given an ISBN-10 or ISBN-13.

I am writing a web-based application in which users (universities) enter the books in their collections into a database. To aid in data entry I am simply asking they input the ISBN-10 or ISBN-13 number and will use a web service/API to fill in the remaining details, if possible. So far I've investigated: Library of Congress, WorldCat, Amazon, Barnes and Noble, Louis (APH), Google ISBN, facultycenter.net, Campusi, and Book Collector.

Which (of the listed or omitted), o' wise MeFites, in your experience is best for such a task?

posted by basicchannel to technology (10 comments total)

Couldn't tell you which one is best, but you didn't mention ISBNdb.com, which has a solid collection of data and appears to have a fairly simple free API.
posted by Partial Law at 2:30 PM on April 17, 2007

IANACataloger, but off the top of my head, I do not recommend relying on Amazon or other commercial sites for this. The information they have is provided by the publisher and is often incomplete or incorrect.

Depending upon how much (and what kind of) information you will be picking up, I'm inclined to say that you should go with WorldCat -- their records are up-to-date and very comprehensive, and are compiled by professional catalogers. LOC records will be equally comprehensive but probably a bit more limited in scope, since they reflect just the LOC's own materials.
posted by Siobhan at 2:42 PM on April 17, 2007

As far as university libraries go, AFAIK the "standard" is WorldCat. I know my university's library catalogues their books using WorldCat---or adds them if they're not there. Then again, if you have access to WorldCat's full database, you'll also already have access to WorldCat's online cataloguing tools.

On a side note, I'm surprised that ISBN numbers suffice if we're talking about university libraries. The number of books from foreign countries or pre-ISBN days must be staggering--hell, even in my room, I have 5 or 10 books without ISBNs.
posted by goingonit at 2:45 PM on April 17, 2007

Thanks for the input so far. goingonit: This is for English-language texts for vision-impaired persons (we are first cataloging the normal texts and requesting accessible versions from the publisher), which is why ISBN will suffice (for now).
posted by basicchannel at 2:48 PM on April 17, 2007

Also, we are providing a manual input form if the ISBN doesn't return a result or they don't have one. This is just to make things easier, where possible.
posted by basicchannel at 2:49 PM on April 17, 2007

ISBN.nu?
posted by joeclark at 3:33 PM on April 17, 2007

Iirc, LibraryThing has an API that may be useful to add to your list.
posted by cmiller at 4:19 PM on April 17, 2007 [1 favorite has favorites]

goingonit: WorldCat was our first choice but they have a verrrry limited amount of data on titles. Their regular search returns a good amount of information, but they restrict what can be retrieved over the web service interface. :( :(
posted by basicchannel at 5:58 PM on April 17, 2007

If this is a nonprofit project, you might want to get in touch with Tim Spalding from LibraryThing. If there is an answer to this question he will know it. He's traveling now -- I just saw him speak at Computers in Libraries -- but I bet he could help you with your general problem as well as your specific problem. He's at tim at librarything dot com, pretty sure.
posted by jessamyn at 10:19 PM on April 17, 2007

You can pull out full MARC records from the Library of Congress by using the Z39.50 protocol.
posted by owen at 4:40 AM on April 22, 2007 [1 favorite has favorites]