User Contributed Data in Libraries
You’ve probably seen by now that OpenWorldCat now supports user data including Reviews and notes. I only expect more user contributed to data to creep into Library related systems. There has been debate over the Wiki-ness of the project but I’m not concerned about taxonomy here. There was a recent post on XML4Lib discussing the WikiD/OpenURL backend. From the very little information presented I think there is a lot more good things to come. The WikiD page over at OCLC sheds some light and looks like there may be some promising web services/data extraction in the future.
I didn’t really understand how much value user annotations could bring until I saw the BBC Audio Annotation project. While still under development the preview gives you an idea of what is possible, including screencasts. People can easily create segments in the audio that can be annotated with information. From the post:
The project we undertook was focused on Annotatable Audio (specifically, but not exclusively, of BBC radio programming) – and we decided to look in an unorthodox direction – towards the possibilities of user-created annotation and metadata. We decided that we wanted to develop an interface that might allow the collective articulation of what a programme or speech or piece of music was about and how it could be divided up and described.
I think of it as an intelligent transcript though much more can be done. I can see it being expanded where people can record their own snippets to be associated with certain sections or even attach other audio files. For example with a speech was given twice but a certain section was changed in later speeches someone could attach the snippet of the change to the original so others could hear what changed and compare.
Our library has a large collection of voice recordings and I can see the ability of users to annotate these with notes, links, transcripts being a very large plus. Odeo already allows people to comment on specific shows, how long before they can comment on specific sections of the stream? On a side note Odeo has their podcast creation section up for many users I think, at least it’s up for me. Very simple and easy to use.
I’ve also seen various posts around the web regarding tags, folksonomies and other user data in the OPAC/ILS. Does user contributed data have a place in the library? Should it be social or personal? How much control should be in the hands of the patron? Right now the OpenWorldCat has very limited things for what a user can contribute but that will likely change over time.
October 28th, 2005 at 9:37 pm
BBC Audio Annotation Project
All I can say is wow. Having the ability to comment on shows at Odeo was nice, but I never thought about the ability to collaboratively annotate audio, including snippets. If you haven’t already go over and read the full post about it, including…
October 29th, 2005 at 8:20 am
The mind races with possibilities in libraries, though one also must compare the BBC’s embrace of the social web with the NY Times step back (with it’s new subscription-only web offerings).
October 31st, 2005 at 5:38 pm
Yes, I think that tagging is going to be huge if someone clever can work out a few details:
1) Ways to use the cataloged metadata you’ve already got to make the tags more powerful, and vice versa (quick example: do a subject browse on “mysteries”. All you’re going to find is a bunch of stuff on medieval miracle plays — probably not what you were looking for. Well, why couldn’t there be a user-contributed subject tag pointing you to the “detective and mystery stories” heading instead? With a user-contributed tag, you can still get the huge benefit of decades of cataloging with controlled vocabularies — all of your mysteries are associated with the “detective and mystery stories” heading — and the benefit of free tagging, which is that it’s very adaptable to how humans actually use language and the fact that connections between concepts evolve faster than our controlled vocabularies do. The two compliment each other. I don’t have to go and tag thousands of mysteries with the “mysteries” tag; I just have to tag one subject heading that all the mysteries are connected to.
This could evolve into huge chains or clouds of different types of controlled and uncontrolled metadata connected together.
2) Ways to make free tags and other user-contributed metadata collaborative and shareable across library systems. When I tag something or write a review or make a “you might also like…” recommendation in my catalog, that same info should show up automagically in hundreds of other peoples’ systems. Amazon.com is a great repository of user-supplied metadata — reviews, rankings, etc. There needs to be an open source, Creative Commons licensed equivalent. Tagging is great, but its power multiplies the more people you’ve got contributing to the system. We need a worldwide system for collaborative metadata.
3) Convincing people in charge that this all is a good idea. This is the main stumbling block, in my opinion. The technical part of it isn’t trivial, but it’s very doable. The human part, I’m not so optimistic about. It will flat out never happen at some libraries (including several of the largest public libraries in the US) because of union contracts that forbid allowing non-library employees to do work for free that a librarian would be paid to do, such as tagging or reviewing materials.
October 31st, 2005 at 9:08 pm
I agree that tagging has some way to go. There will likely be quite a few failures before a good hybrid is created, especially in the realm of structured data like a library. I’d be interested in seeing if Amazon tackles such a feat.
For the second point I’ve also wanted such a service. Amazon allows pulling of information though the inability to cache it is a killer. It would be really nice to see Amazon partner with libraries to provide data. I had high hopes for the OCLC offering but question if it will be any more open then any other review service. Right now things are becoming even more fragmented as more and more services roll out. I currently use listal for my collection as it’s very easy to get my data out (rss for everything and xml upcoming). It also has some amazon integration with it.
As for convincing I think the main hurdle will be a few libraries making great examples. Once there are some examples to point to I think it will become easier (as in any field). There will always be some that don’t flow and then there is the slew of privacy and other issues.
October 31st, 2005 at 10:31 pm
Should have read my aggregator before commenting. You may find the link below interesting. It discusses authority and tagging (synonyms).
http://atomiq.org/archives/2005/10/tagging_tags_to_make_synonyms.html