This was a four hour program, and even though I did not take my best notes ever, this is still going to be a long post… Also, full disclosure – I’m a total Tim Spalding fangirl.
Robert Wolven moderated and began with a great introduction covering the gamut of opinions. The speakers included Roy Tennet, Jennifer Bowen, Diane Hillman, Martha Yee, and Tim Spalding. See the program description for their full credentials. Tim has his slides up and links off to Diane’s also.
Robert explained that the format will consist of trying to pair up slanted views then allow for discussion and that these talks are to start discussion, not give answers. The program is not about predicting the future but creating it. He expects we’ll go away knowing less about the future but understanding more.
Roy Tennet went first. He talked about library catalogs at the network level. Then – users built workflow around libraries. Now – library must build services around user workflow. Discovery happens elsewhere. He had a great image of the internet as a street map of a big city including major shopping centers. He talked about concentration and diffusion. In the old way – added value was only for your users. He talked about WorldCat – a union catalog would allow the added value to be for everyone. With concentration/aggregation you can have search results ranking by holdings data, usage data. You can have recommendations of related books, user contributed content can be shared, and user contributed content can be made better.
Roy talked about WorldCat identities and the fact that OCLC is big enough to cut a contract with Google, it allows small libraries to play in big spaces. He also talked about putting special collections into Wikipedia (University of Washington Libraries Digital Collections – salmon example).
Concentration allows for a web scale presence and mobilization of data. Diffusion allows disclosure of links and services.
Next up was Jennifer Bowen talking about eXtensible Cataloging. She just had an article come out. She said that Roy talks about network level, she comes from a different perspective, but it’s complimentary. She plans to mention some provocative things and move on. She’s working on designing a set of open source tools that facilitate resource discovery and metadata management. They’re planning to release July 2009.
The tools will empower libraries to customize/develop discovery solutions (without having anyone with coding skills on staff). They will integrate library content into various web environments, enable sharing of metadata and software, and expand the role of libraries. While the tools will be compatible with an ILS (via OAI-PMH), they are not an ILS itself.
There will be out of the box user-interface functionality (but that’s not the main point). There will be toolsets for customization, web application development, and metadata enrichment. They’re building an infrastructure for moving beyond MARC. The XC network – point – metadata is going to move all over the place. Some of the user interface apps will allow metadata back into the system, enabling customization and development, empowering libraries. This “lowers the bar” for local level development.
XC works alongside network level applications. This allows libraries to focus attention upon needs of local users. When you have out of the box tools that you can’t customize, you don’t focus on users needs because you can’t provide it. XC also encourages user research and new roles for catalogers/cataloging, including designing local applications and engaging in user research.
Some of the challenges include keeping an open mind, presenting options, not objections, thinking broadly, and integrating library content into web environments.
Bring metadata to the users. Create new “measures of success” for metadata standards. Compatibility with web standards is now essential. Existing metadata might not be compatible.
Other challenges for catalogers include perceived loss of control and rethinking standards development.
- make legacy metadata as usable as possible
- XC metadata services
- authority control matching
- share post-MARC metadata
- share user-generated metadata
- share metadata enrichment services
- expanding the role of libraries
- uncover and address evolving user needs
- facilitating the goals of the researcher
- participating in learning management system environments
- exploring new roles for catalogers
- design and test metadata services
- manage flow of metadata
- system/application design
- user research
During the comment section Tim said that he agrees with Roy mostly, but takes exception with his claim to web scale. If you add catster and dogster the two combined have the same traffic as WorldCat, so it’s not WorldCat that is so big it can make a deal, it’s librarians.
[last speaker] OCLC is not the deathstar. It has enabled us to do so well with datasharing, but in order to be able to innovate the data has to be able to move around freely. That’s scary for people whose business model relies on data paying tolls as it moves around. The data has to be able to be open for use and reuse. The business model has to be services.
[Roy] Certainly these issues are not new to us. The data aggregation at WorldCat is something you all have built, thousand of people. How can we use that to the best of our member libraries? If you want to toss it out there and let Google and Tim have at it, then that’s what you need to be telling us. [Ellie aside – let’s all write OCLC and let them know, yes we want to share data freely. With the open library project too.]
[Bob] Asked about scope of what we want to present to people since “can’t have” is a relative term.
[Roy] People are interested in what they have to go through to get something. We need to be better about how we present that to people. He feels they’ve achieved that to some degree with WorldCat Local. We need to do that last mile in a more effective way – up to including a link to buy this used from Amazon.
[Jennifer] Use facets – like I want this right now, my paper is due in 24 hours. Being able to look at yes this is on the shelf and yes you have permission to borrow it and yes the library is open until 2am. [Ellie – very interesting idea to get to that level of granularity within the catalog!]
[Bob] From Roy’s perspective there is no local user – everyone is on the web.
[Tim] Centralization has been squashing local. Dewey killed Cutter. Now we’re getting into a phase where that starts to come back, where local info can come back and not hurt everyone else.
The third speaker was Martha Yee with “What I have found out from an attempt to build an RDF model on FRBR-ized cataloging rules.”
She said we spend entirely too much time doing clerical work to keep individual catalogs under control. She went fast and furious and I gave up on keeping up with my notes so I could keep up with just following along.
She started with some definitions. The semantic web is a way to represent knowledge. The vision is that the web could be a shared database, hyperdata replacing hypertext.
This is an experiment. Then she got into detailed cataloging stuff that was very interesting, but that was too out of my realm for me to summarize.
She made a distinction between granularity and complexity. And MARC has some complexity we don’t need. We need our granularity to be more interoperable. Granularity isn’t bad, but confusion about what something is, is bad.
At this point we had a small break. I got myself to an outlet and started recharging.
Next up was Tim. Again his slides are available on his site. Warning – library science being practiced without a degree! He said there are many competitors, but none of them care about you (librarians). He showed some famous people libraries (e.g. Thomas Jefferson) that have been cataloged in LibraryThing and explained they are all work done by LibraryThing members in their spare time.
He showed the section “common knowledge” where members can catalog stuff that isn’t typically in records, like characters. He showed the Star Wars series and related series. This page assembles more info about the Star Wars series than ever assembled. He showed the entry for Huck Finn, scrolled through hundreds of editions and explained that regular users combined them. He showed the paranormal romance tag, cyberpunk, cozy mysteries and explained tagmash and how it has some of the power of hierarchy (but not all). He explained that relevancy is built into the system.
The tag war is over. Tags aren’t better, but anyone who says they aren’t useful hasn’t looked at the evidence, if what you care about is finding things rather than asserting ontological reality.
Tim talked about the physical basis of classification – a book has 3-6 subjects – no – that’s how many fit on a card. On the card (and in our catalogs) subjects are equally true, subjects never change, only librarians get to add subjects.
The world is not hierarchical, ideas are much more subtle and complex. There are two futures. In one the world ends – you are paid less, programmers still get paid. In the other you move up the stack.
Concluding tangent – a new shelf order, replaces dewey, free (open source), modern, humble, decided socially, level by level, tested against the world, assignment is distributed, I write the code, you be the Jimmy Wales.
There was a question from the floor that I think was trying to make the argument that since librarians have to know HTML they should get paid like programmers. The only comment I made was “oof!” but I now feel like adding something along the lines of that’s like saying kids playing with tinker toys should get paid like construction foremen. I hope I just misunderstood the question.
Next up was Diane Hillman with “A Has-been Cataloger Looks at What Cataloging Will Be.”
The extinction model is one of our choices, the retooling model is the other.
Extinction – Why should we change? Make them change! The sky is falling!
Retooling – What do we need to do in this new very different model? Retraining.
New job title – metadata librarian
We need to look for tall buildings to jump over or our jumping skills will atrophy and we’ll fall into a hole. We need to design data to play well with others. Working with data in aggregate, not just one item at a time. “The day of the ILS, if not dead, is on life support and I for one am not filling the bag there.” You want them to get to your content whether they use your portal or not. Dis-integrated library system. The discovery piece is the first bit coming unglued. Most vendors are accepting that people are using ILS for backend but choosing different front ends. We need to stop commoditizing data. Without open data we cannot compete with data providers.
[Roy] Glad to hear a call to Tim to give his data away. Mention of WorldCat API.
comment on Tim’s presentation – user generated FRBR-ization
Have to be able to recognize and catalog a point of view
[Tim] Lots of web people don’t think the semantic web is the answer, but librarians have really latched on to it.