Remixing the library

A talk given by Jon Udell at the Global Research Library summit, October 2007.

Abstract: In an online world of small pieces loosely joined, librarians are among the most well qualified and highly motivated joiners of those pieces. Library patrons, meanwhile, are in transition. Once mainly consumers of information, they are now, on the two-way web, becoming producers too. Can libraries function not only as centers of consumption, but also as centers of production?

About five years ago I started an online project that I called LibraryLookup. In its original and still most widely-used incarnation, it's a kind of mashup that connects a web page about a book, at Amazon or Barnes and Noble or elsewhere, to the database record in a local library's catalog that says whether the library holds the book, and if so whether the book is available.

Of course it was always possible to read about a book at Amazon or Barnes and Noble, and then, with some judicious cutting and pasting, to capture the book's title or ISBN, switch to the library's catalog, paste in that information, and search for the book. But those two resources -- the book page and the OPAC record -- were widely separated in information space. LibraryLookup shortens the distance between them, and reduces the lookup to a single click.

In the science fiction novel Dune, a race of strange creatures called guild navigators plays a crucial role in the galactic ecosystem. The guild navigators are the ones who make space travel possible, and they do it by "folding space" so that distant locations are brought together.

Although software people and library people belong to different tribes, we're all guild navigators in this sense. When we integrate software and services and information resources, we're folding information space. I guess that's why, although I'm not a card-carrying member of the library tribe, I have a sort of adoptive status, and it's why I'm here talking to you today.

Since 2002, this LibraryLookup idea has spread slowly and steadily. It's no mystery why. If the book you're reading about on Amazon or Barnes and Noble is available in your library, you can read it for free -- and sooner, probably, than a retailer could ship it to you.

I've done a bunch of variations on the LibraryLookup theme. Several of them take advantage of a great little service provided by OCLC called xISBN, which expands an ISBN to the set of ISBNs representing a work, and makes an ISBN-based lookup much more reliable.

Here's my favorite variation. When I see a book that interests me, I toss it onto my Amazon wishlist.

I have a little service that scans Amazon every day, picks up the ISBN for each of the books on my wishlist, runs each ISBN through xISBN to expand out to the set of ISBNs that might appear in the library's catalog, and checks each for availability. If any of the books on your wishlist has become available in the library, I get an RSS alert.

Over the years, the list of OPAC [online public access catalog] systems supported by LibraryLookup has grown to about twenty, as I've identified query patterns for these different systems. Or actually, I should say, as librarians and patrons have helped me to identify those patterns.

I particularly remember one email exchange with a librarian who asked why her OPAC was not supported by LibraryLookup. There are a couple of reasons why that can happen, but it all boils down to whether or not the OPAC exposes a URL that can be used to search the catalog. In the end we concluded that her OPAC didn't, at which point she answered her own question in the following way:

"Oh, I see. We bought the wrong kind of software."

What she meant by "the wrong kind of software" was: a system that's not open to lightweight, spontaneous, opportunistic integration.

And I had to agree. Her library really had bought the wrong kind of software.

I'll bet she figured that out before the IT person who had gone through the evaluation and purchasing cycle. From the IT perspective, the system they bought met all the requirements at the lowest cost. Lightweight, spontaneous, and opportunistic integration just wasn't a requirement. An IT person wouldn't necessarily be the first one to appreciate that requirement. A librarian would, and as I've learned firsthand, they do. I hear from a lot of librarians, and they're really clued in to how this kind of integration can work.

Here's one of the librarians I've turned into a URL hacker:

From: Janet Lefkovitz
To: "Jon Udell"

Hi Jon- The bookmarklet generator is working for Aleph500! I've dragged the ULI button to my links bar, and have used it already. One small change: change the Aleph to Aleph500, and state somewhere that it's compatible with versions 14 and up. Many thanks.

I like to encourage and reward these efforts, so I credited her with the discovery:

At this level of the stack, integration winds up being a game of capturing, analyzing, transforming, and interconnecting URLS, and by extension the resources those URLs point to. What I've discovered is that although a lot of librarians have never played this game before, they like playing it with me -- and they're good at it.

This leads me to the following observation. In an online world of small pieces loosely joined, librarians are among the most well qualified and highly motivated joiners of those pieces.

Now I am not directly involved with the kinds of research libraries that most of you are concerned with. My perspective is shaped largely by my experience with two interlinked libraries in the small town where I live. One is the public library, the other is a college library. So you'll have to take my conclusions with a grain of salt. But I hope they can generalize.

So, what is the environment in which this joining of small pieces is taking place? Well, it's evolving, and it includes libraries as they begin to embrace the remix culture, and to provide formal ways to do the kinds of things that I'm doing opportunistically with LibraryLookup.

One of the most interesting places where you can see this happening is in Ann Arbor, Michigan, where a couple of guys are working together in a really interesting way. One of them is Ed Vielmetti, who blogs as Superpatron.

As a superpatron, Ed likes to extend the capabilities of the Ann Arbor District Library and its online catalog. Here's an example where he mashes up the Internet Movie Database with the video holdings of the library -- it's basically LibraryLookup for DVDs.

At the same time that Ed's experiments are pushing the envelope in terms of what the library's systems can already be made to do, there's also a superlibrarian, John Blyberg. He's a technologist who works for the library, and in a wonderfully complementary fashion he's been observing the kinds of things that Ed is doing, and has been adapting the library's OPAC system to make it friendlier to remixing.

Among other things, John has defined an XML-oriented interface to the system -- it's called PatREST, for patron REST, where REST stands for Representational State Transfer, but for our purposes here just means web friendly -- and he's implemented that layer on top of the library's catalog system. So you've got Ed exploring the possibility space, and John working to enlarge that space, and together they've created a virtuous cycle of innovation.

Now this is obviously an extreme example. You are not going to find a superpatron of Ed's caliber and a superlibrarian of John's caliber in every town. But I think the dynamic at work there can apply more broadly. And if it does, it will matter that these patrons and librarians are situated in a local context.

One the one hand, the library is global in its reach, aggregating growing numbers of subscription resources from around the web. On the other hand, it's chartered to be a repository for certain kinds of local information.

You could argue that the local part of the library's function will shrink as the global part grows. But there's a counterargument: the web 2.0 architecture of participation.

Several of my own personal projects explore local uses of the web to gather and organize community information resources. I have a hunch that local libraries can play an interesting role there, and I'll give you an example that points toward the kind of thing I'm imagining.

I live in Keene, a town of 25000 souls in the southwest corner of New Hampshire. Our nickname is Elm City. There are lots of Elm Cities, but in January I registered the name elmcity.info and launched a little experiment in the aggregation of community-based information.

Elmcity.info is entirely dedicated to the proposition that, through syndication and tagging alone, sources of information about the community, as they begin to emerge on the web, can be usefully aggregated and made available.

Here's the photo gallery. Currently it combines two sources. In the left column there's a feed of photos about Keene from Flickr. (Yes, that is Barack Obama. In New Hampshire, during political seasons, you can't turn around without tripping over a politician.) In the right column there's a feed from Lorianne DiSabato, a local place blogger who writes a wonderful daily photo essay about life in our town.

By combining these feeds I'm trying to show that if you want to contribute to a stream of Keene-related images, you needn't think in terms of joining and contributing to a Keene-specific repository. Instead you can -- and probably should -- think in terms of using existing photo-related services out on the web, and tying things together by means of syndication and tagging.

So here I'm inviting folks to use Flickr to tag photos that they'd like to appear in the stream, but by also sourcing photos from Lorianne's blog I'm trying to show that elmcity.info can syndicate from any photo source.

In a similar vein, I'm endlessly frustrated by the failure of people in my community to latch onto any of the available online event services. So as a demonstration, I've here woven together the web calendars from the four or five major organizations in town that produce web calendars -- including the newspaper, the college, and the chamber of commerce -- and I've synthesized a common stream of events.

This was one of those thankless HTML screenscraping chores which I'll probably find myself doing until the end of my days. But again the point is to show that there's a better way. So one the sources is Eventful.com, a leading event service. But it could equally be Upcoming.org, or a combination of the two, or others as well. What matters is not which service you use, but that you can syndicate information from that service.

There's also a community-based version of LibraryLookup. In an effort to broaden its appeal, I did away with the RSS feed and switched to email as the common denominator for notification. Unfortunately this is still a bit too geeky for mainstream use, and so far it's gotten almost no traction in my community. To those of us living in the web 2.0 bubble, it's easy to understand that something I post here with a certain tag can appear over there. But to the vast majority of folks living outside that bubble don't get that. I think libraries can help expand the bubble, and I'd love to see them try.

Now arguably these kinds of information services -- the photo stream, the event stream -- belong more properly to the newspaper than to the library. But I have another example that may come closer to the mission of a local library. I've been on a campaign for a while now to uncover sources of public data -- for example, crime data. There's been a debate in my little town about whether or not we've been having a crime wave over the last couple of years, and the debate has been raging in a vacuum because nobody has any data to support the arguments they're making.

So I went looking and found a couple of things. First of all, it turns out that the Department of Justice has a certain amount of crime data for the nation, and also state by state and town by town. That data is available at the DOJ website, so I've latched onto it.

Separately I've lobbied the police department to let me have the spreadsheet containing the crime reports that they've recorded locally.

It took a while, and I finally had to hire a lawyer to get them to release this information. But I think that once I got hold of the data, and showed them how to geocode it (which is a capability they hadn't had before, they had no way to show incidents on a map or analyze them geographically), and showed them how the taxonomy they're using on the ground doesn't line up with the taxonomy that their reporting system is using when it feeds data back to the FBI, and showed them how services like Many Eyes can enable collaborative analysis and visualization of the data.

I think all this has broken the ice, and that they'll be willing to work with me on this stuff going forward. At least I hope so.

This example is just a tiny part of a vast movement toward transparent operation of government at all levels. On the local level, here's an interesting question. What institution could be chartered not only as the repository for this information, but also as a partner in the design and development of the architecture of the information?

I doubt any local library currently thinks it is chartered to help people organize the homegrown information and data they are increasingly producing. But to me that seems like a valid and plausible mission for a local library. I can imagine the library functioning not only as a repository and aggregator of existing sources of information, but also as an adviser, a consultant, or a coach to local organizations as they create new information sources and build out new information services central to the life of the community.

Police departments and school boards and downtown retailers and homeowners have no special skills in, or inclination toward, information architecture or taxonomy or service composition. Libraries and librarians do.

Meanwhile I can't help but notice that one of the things public libraries do a lot of nowadays is computer training. I often go into the library and see one of the librarians sitting with someone -- usually an older person -- and walking them through the basics of how to use Windows and Office and the Internet. I wonder if there's not an interesting role here for the library to also advise organizations and individuals on how best to shape and organize the information that those organizations and individuals will be creating and publishing themselves -- information that will represent their public agendas in the community.

I want to finish up with a thought about the library as a physical place. In my town, and I think in a lot of towns, the library is one of the only public spaces where you can gather together with other people, use WiFi, and not have to buy a steady stream of four dollar lattes in order to justify your presence in that environment. As a home-based employee I sometimes take a break from the home office and set up a mobile office at the library, where I get a lot of good work done. But it has to be quiet and solitary work. Things get awkward when I need to do something collaborative. And it's not just me. Over the last couple of years I've noticed a growing number of mobile professionals who will drop into the library at various times during the day to use it in the same way I do.

Sometimes these folks need to talk on their cellphones, and that's a problem. Sometimes they need to talk to one another, and that's also a problem. So I have in mind that one part of the future role for libraries, especially public libraries, could be to create a public space where it's not only access to the Internet that's provided, but also an environment in which people can use that shared Internet connection in a collaborative way that celebrates their togetherness in that physical place.

One nice example I've seen, since I've been making trips to Redmond, is the Crossroads Mall down the road from here. There's a branch of the public library situated right in the center of the mall, where it really feels like part of the public forum or bazaar.

The patrons of that library, like the patrons of all libraries, are in transition. For a couple of generations, they've been mainly consumers of information. Now, on the two-way web, they're becoming producers too. Modes of production that have been dormant since the dawn of television -- like storytelling, and letter-writing -- are becoming active again. I hope we'll see the libraries of 2020 functioning not only as centers of consumption, but also as centers of production.