If you ever forget how to pronounce the name of the company “Zepheira,” just remember that you have nothing to “fear” from its various initiatives. The name comes from Vakhtang Sirunyan’s Zepheira, a favorite painting of Eric Miller’s. To him, the name is the perfect way to symbolize the company’s focus on “the art of data.”
Miller, Zepheira’s president and founder, had previously led the World Wide Web Consortium’s (W3C) semantic web initiative at the Massachusetts Institute of Technology (MIT). In 2007, he wrote a blog post to synthesize the new direction Zepheira was taking. It is “a professional services turned product company that spun out of MIT in 2007 to … ‘help organizations use the Web to connect, visualize, analyze and augment data assets across system boundaries.’” Today, Zepheira’s main goal is to help libraries gain more visibility on the web through the use of linked data.
“We’re not a gigantic organization that takes ourselves too seriously—we’re professionals that enjoy what we do,” says Jeff Penka, Zepheira’s VP of product management. “A number of the folks that are the core members of Zepheira came together about 2002. As a team, many had played significant roles in structured semantic data, creating vocabularies, ontologies, the underlying standards that are the web that we use today, and the web of data.”
Sometimes open standards are branded for private industries, and people forget that they were meant to be used by everyone, he says. In Zepheira’s early days, its team members used emerging web and data technology standards to show customers how to unlock hidden values of data by moving it from silos into structured forms. They leveraged semantic and web technology to help companies create new products or to do prototype work. “It was across a lot of different industries,” including pharmaceutical and finance, “but the common thread was the need for the understanding of how to better use data,” says Penka.
Eventually, Zepheira started working on unlocking data to reach new audiences, not only to attract new customers, but also to correct a problem the team was seeing: Not everyone’s data was visible on the web. This was especially true for libraries and other cultural heritage institutions. “And given all of our history with the web and how much we also love libraries, in all their forms and what they can provide … there’s a real opportunity here to really change the face of the web,” says Penka. He compares the current state of search to ready reference in libraries: When internet users do a search, they are accustomed to seeing a box on the right side of the results screen (sourced from a knowledge graph) that sums up the most popular result on the list. “It’s the stuff I want to get a quick answer to. But I don’t necessarily want to go to a webpage, and so the data that’s now available is allowing us to get those types of answers,” he says. “Yes, we have webpages and websites, but increasingly, the data itself is able to be understood, and something’s able to happen with it.”
For libraries, this could mean opportunities to post hours and locations or to promote upcoming events. Adding library data to the web also makes the library’s catalog, special or digital collections, and e-resources available to a bigger audience, as well as its job search or tax assistance programs, etc. Academic libraries could advertise student, faculty, and alumni services. “[O]ur data as an industry within libraries is very rich behind the walls but not really understood out in the web. And so we really looked at that as a passion point for us over the last 2 or 3 years after the work that we did with the Library of Congress, as the technical architects of BIBFRAME,” says Penka.
Libraries use MARC to describe resources such as books, music, and DVDs. Penka says the Library of Congress (LC) realized that “MARC has really reached its capacity. We can keep trying to make MARC better, but we need something else. And so they initiated a project at the Library of Congress to [ask], what would the future of geographic and library description look like so that the web could actually use it, as well as a broader networked world?” says Penka.
Thanks to previous collaborations, the LC was quick to ask Zepheira to come on as a technology partner for the BIBFRAME vocabulary, its bibliographic framework initiative. Penka says the organizations agreed that BIBFRAME would need to be an extendable framework based on linked data, because libraries have other descriptive standards besides MARC. If it could include other standards, it “could act as a unifying vocabulary for the whole industry … so libraries could basically have a linked data-driven vocabulary that describes everything they need as richly as they need it.”
BIBFRAME leverages the W3C’s RDF (resource description framework) model—a standard form of communication for data on the web—to turn descriptions into linked data that can be made available to the web in many different vocabularies. “We’re putting it into BIBFRAME, but if the audience that we want to talk to understands Schema.org, because we’re natively in linked data, we can actually crosswalk over into Schema.org very quickly,” he says. “It was good on LC to be able to recognize that MARC was great inside of our systems, but we were limited in our ability to work across other systems.”
Penka says working on BIBFRAME has paved the way for other Zepheira activities such as the Libhub Initiative and the Library.Link Network.
The Libhub Initiative
The Libhub Initiative, a community focused on learning about visibility and linked data, grew out of Zepheira’s workshops for libraries and industry vendors on how to use BIBFRAME. “[W]hat we found was there was a lot of questions about the practical use of BIBFRAME, and there were a lot of misunderstandings. And so we started to have more discussion, and then people heard about this workshop that we did. … And so what that evolved into is the linked data and BIBFRAME training curriculum that we introduced,” says Penka.