Tuesday, December 20, 2011 |
The days of the card catalogue are long over, but there are even more innovations ahead in how the information available at libraries is organized. In a recent interview on Spark, David Weinberger, author and co-director of the Harvard Library Innovation Lab, talked to host Nora Young about ShelfLife and LibraryCloud, two of the lab's ongoing projects that will transform how we use libraries.
ShelfLife is a new way to browse and discover library resources online that adds a contextual dimension, which is shown "by using a very traditional visual representation: a shelf. "
The image of the shelf is turned sideways so the book titles on the spines can be read, and users can scroll the shelf. "This is in some ways a very old-fashioned way of seeing a work in context. The idea is that works are always in a web of other works, they have meaning and significance because of the connections they have to other works," Weinberger said.
But ShelfLife also adds a couple of features that are not old fashioned in the least. "The first is that you can always pivot on any particular book," Weinberger said. "So you click on a book and it shows you all the categories under which it has been classified by librarians but also by users in various ways, by tags and the like. And you click on one of those classifications and it redraws so that you see all the works that are within that classification, including the original one that you're looking at."
The other new feature is that the colour of the book's spine is mapped to the book's popularity. "We give weightings to the various ways in which users can have used the work. So you'll be looking at the shelf and you'll see not only a set of works that are like it in some way...but you also see, by the depth of the colour blue, how relevant that work has been to your community."
Weinberger stresses the advantages of having such contextual information. Simply by putting it onto a stack of books that are like it in some way, "you're seeing it in a context that other people think it makes sense in."
How will that change our relationship with libraries and how we access information? Weinberger says that it makes available information that libraries have, but haven't made use of. Weinberger cites circulation records as an example. "In the university library we know when a book has been used in a class or put on reserve...or while it was out, did somebody call it back in," he said. "It turns out to be a pretty good indicator of how relevant the work is at that time."
Weinberger sees value in collecting as much information as possible about the usage of works, because that adds to the contextual richness.
"How your social network, the people that you know, or in your community, understand or value a work can be...a tremendously relevant indicator of how important or meaningful it's going to be to you," he said. "So there's no reason why, in a digital environment, we shouldn't collect as much of that information as the privacy laws allow, and local norms allow, and use that in anonymized form in order to help see the shape of the usage of the library."
LibraryCloud is a collaborative project that is a way of gathering and making available a range of useful data. "There always has been, in libraries, a tremendous amount of information, but in physical libraries much of it has escaped because there hasn't been any way to record it," Weinberger explained. "The aim is to gather circulation data, user reviews, acquisition records, whatever information physical libraries have about their works that they're willing to share, bring it all into one digital spot and make all of that information available to developers who may want to do interesting things with it." As an example, he described a developer using the data from a group of libraries to compare patterns of usage at "public libraries versus university libraries, or public libraries in rich areas and public libraries in poor areas."
Weinberger listed several other applications that LibraryCloud lends itself to, including highly personalized recommendation engines and library analytics. "I think one important application is to explore what's often called 'the long tail' of works," Weinberger said. "[At] most libraries, there's a relative handful of works that get checked out. And then works sort of fall into disuse, people forget about them. Something like LibraryCloud can be used to try to identify items that are in this long tail of semi-forgotten works and make them interesting to people, so they'll pay attention to them again, get value from them."
Weinberger cited another major advantage of digitization's ability to gather and organize data in a myriad of ways. "With physical collections, very smart people have to make decisions about what they anticipate will be the most desirable and interesting set of works to buy and put in place, and how to organize them in the single best way. But there isn't really a single best way," Weinberger pointed out. "People are interested in things that we cannot possibly anticipate, and there are ways of thinking their way through works, to navigate and find, we also cannot really fully anticipate."
Weinberger has written extensively on information and how it's organized. His new book, Too Big to Know, about the changing concept of knowledge, is forthcoming in January. When asked how his work on libraries fits in with that, he responded: "There's a deep relationship between how we organize works, and what it means to know them. And also, what we think they are. Those three questions are all deeply entwined in ways that we didn't always recognize, I think. Now that we're able to separate those ...we can have multiple playlists so to speak, of knowledge over the same set of information. We lose the sense that we traditionally in the West have had, that is to know something is to see its place in the universe, its single place in the universe."
Weinberger acknowledged that this new way of looking at knowledge has its drawbacks. "We've organized ourselves as cultures, to a large degree, around what we agree we know. And when you have multiple ways of knowing, multiple ways of organizing, the society loses one of its deepest organizational principles," he pointed out. It also gets more difficult to dismiss some points of view, even if they're wrong — "it gets harder to rule them out when every bad idea can be made globally public through the web. "
Nevertheless, Weinberger believes that the digitization of information offers tremendous potential: "I'm still very optimistic about it."
Spark will also feature David Weinberger in a long-form interview about his new book, Too Big to Know, on Jan. 8.