If you had the word “iterate” on your PLA conference bingo card, this was your year. Concepts such as rapid prototyping, failing quickly, and agile development made their way into many of the program presentations, and all the featured speakers addressed issues related to embracing constant change. With that in mind, I’d like to take a second pass at a topic I covered in this column a year ago: how can libraries make better sense of the opportunities afforded by the open data movement?
In case you haven’t been committing my writing to memory all this time, “big data” refers to the significant volumes of information generated by the users of a given organization or service. In the public sphere, we are seeing governments opening up these datasets to the public, allowing the civic-minded to create new applications and interpretations of this material.
This all lines up quite nicely with current library trends. It could be a next step for those working with STEM (Science, Technology, Engineering, Math) students. It can connect to digital literacy efforts, offering insight to the code-curious and practical applications for those looking for some direction. And it could create a very real platform for civic participation, opening the partnership door for groups like Code for America. For libraries, embracing open data principles can allow us to revisit our own metrics and create new tools for improving public
services. As I wrote a year ago, “there’s something out there. We’ve just got to find a way to harness it.”1
Best practices for open data in libraries are still taking shape. But we’re starting to gain some toeholds—not just for public data practices, but in examining the way we use data to measure and provide service.
Libraries as Open Data Hubs
Perhaps the easiest way to get a foot in the door with large datasets is simply to think of them as another item in the collection. Even if we are not entirely sure what we want to do with this information, we can create a space where all stakeholders can start learning together. Anastasia Diamond-Ortiz, knowledge manager at the Cleveland (Ohio) Public Library, is a strong advocate for this approach. Here are her thoughts from an interview I conducted over email.2
Public Libraries: How do you see libraries using open data?
Diamond-Ortiz: Libraries have the ability to see beyond the single use for information, including open data. That is what is most exciting. From hosting an edit-a-thon of OpenStreetMap (www.openstreetmap.org), which simply requires space, to devoting technology staffing and resources to hosting local civic data, there is an entry point for everyone. Public libraries are civic institutions that are seen as positive and open to all views. We absolutely need to embrace this and start participating in the conversations happening around us.
PL: What’s the best way for libraries that aren’t familiar with civic hacking to get oriented?
Diamond-Ortiz: The best and most accessible way is to start looking at the vast array of maps and charts being produced by cities and other interested groups. The work of Chicago’s Open City and MIT’s Civic Data Design Lab are two good examples that are very visually appealing. Seeing some of the possibilities allows libraries to start thinking about questions instead of worrying about the answers.
PL: How can we make this stuff interesting to the lay patron? What’s the lure beyond just the hacker stuff, which might be intimidating to many?
Diamond-Ortiz: Public librarians see a broad cross-section of the community walking in the doors every day. Civic data is about coding and data, of course, but it is also about listening to our patrons and engaging them in the conversations. Besides setting patrons up as informed consumers of data, we also need to consider how and why patrons might want to contribute to datasets that have the potential to improve the quality of life for everyone. This should not be a one-way relationship.
PL: What can libraries present as a dataset of their own?
Diamond-Ortiz: As an example, a library could contribute aggregated technology usage data. What if someone took those data and layered them with broadband subscribers in a neighborhood? What if the local government used the resulting layers as a planning tool when thinking
about where to locate wireless hotspots in a city? Not knowing how and where data will be used is part of the fun.
Data, Privacy, and Library Service
While there is plenty of opportunity in libraries creating open datasets, there are plenty of reasons to be cautious. Any amount of data a library makes public needs to be weighed in light of its risk to user privacy.
Does this hold us back? In the age of customized everything, does treating the information we share on our users come at the expense of better service? To test this theory, Darien (Conn.) Library (DL) recently made the decision to make its users’ reading history opt-out. I spoke with Amanda Goodman, DL’s user experience librarian, and John Blyberg, DL’s assistant director of innovation and user experience.3
PL: Talk about the process that led to making patrons’ reading history opt-out.
Goodman: Businesses have the advantage on libraries because they have so much data about their customers. They can therefore create personalized interfaces, shopping experiences, and more to entice repeat visits and spending. Libraries, on the other hand, have always valued patron privacy and therefore not kept any records [beyond] the bare minimum. In order to meet patron expectations for personalized reading recommendations and meet the frequent request for reading history, Darien Library’s staff met to discuss keeping patrons’ reading histories. Once we agreed that we would like to make this feature available, we discussed how to communicate the change, user privacy, and whether this should be an opt-in or opt-out process. Opt-out was decided as the default.
Blyberg: I would add that nobody takes privacy more seriously than libraries and we’re no exception and given the type of personal information most people willingly offer up to corporate websites like Facebook, we feel that making checkout history opt-out is a natural step to take.
PL: What kind of response did you receive in the lead-up to the big switch?
Goodman: Patrons were often surprised that this was not a feature previously available. We had around a dozen people opt-out of the system before launch on January 1, 2014, and very few since then.
PL: Did you receive any unexpected feedback after you made reading history opt-out? Has this had an effect in how patrons interact with library staff?
Goodman: I am not aware of any unexpected feedback. As for interacting, I had a patron yesterday that asked us to pull up her history to see if she had read a book before. I showed her that we only had her history since the beginning of the year. However, my colleague stepped in and reminded the patron that she had came to the book discussion for that book and she had in fact read it last summer!
PL: What’s the next step in how you plan to interact with data tied to patron behavior and library use?
Goodman: The next step is to look at models that use patron reading history to create personalized recommendation engines similar to GoodReads, LibraryThing, and NoveList. If such an implementation was built, the data would be anonymized and people would see recommendations based in whole upon what the group read next.
Better Libraries through Data
Embracing more open practices with data could be a catalyst for reexamining privacy in libraries. As we continue to create new models of transparency and fluidity in our services, it will become all the more essential to examine these key elements of our profession.
- R. Toby Greenwalt, “In Search of Better Metrics,” Public Libraries 52 no. 2 (Mar./Apr. 2013), 22.
- Anastasia Diamond-Ortiz, email interview with the author, Apr. 7, 2014.
- Amanda Goodman and John Blyberg, email interview with the author, Apr. 4, 2014.