Curating Digital Collections to Go Mobile (Category: Research, Creative / Artistic, Entrepreneurial)

Name of Submitter(s): Mitchell Davis
Name of Team: BiblioLabs
Organisation: BiblioLabs, LLC

As a direct result of its collaborative work with the British Library, BiblioLabs has developed BiblioBoard, an award-winning e-Content delivery platform, and online curatorial and multimedia publishing tools to support it. These tools make it simple for subject area experts to create visually stunning multi-media exhibits for the web and mobile devices without any technical expertise. The curatorial output is almost instantly available via a fully responsive web site as well as through native apps for mobile devices. This unified interface incorporates viewers for PDF, ePub, images, documents, video and audio files allowing users to immerse themselves in the content without having to link out to other sites to view disparate media formats.
To date, over 250,000 people have used the British Library 19th Century Collection app for the iPad, which was the impetus for BiblioBoard. Released in conjunction with Apple’s Worldwide Developer Conference in 2011, the app has been downloaded in almost every country and it has logged close to 700,000 sessions. On its launch, the Sydney Morning Herald raved
“…libraries will become places, both physical and virtual, where the public accesses information and creates meaning in a purely digital realm. What might this look like?....the amazing new app the British Library has just put out.”
The enthusiastic responses from both the press and the public to the British Library 19th Century Collection app lead our team to explore 2 research questions:
• How can we make it affordable for any institution to create a similar experience, specifically on mobile devices, with their digitised content?
• How can we facilitate the curation and organization of content to enhance its discovery, especially when it is unlikely the end-user will be searching for a known item?
The resulting product, BiblioBoard, with its publishing suite BiblioBoard Creator, was launched in 2013 and prominently features content from the British Library. National libraries, the National Archives (UK), museums, universities and others have used BiblioBoard to curate and provide access to their digitized collections. Schools and universities, including Liverpool and Sussex, are taking it a step further and using BiblioBoard to create and host open educational resources. On the other end of the spectrum, some museums and libraries, including the British Library, are using BiblioBoard to develop and distribute revenue-generating products.
URL for Entry:; (britishlibrary/welcome1)

Job Title: Chief Business Officer

Background of Submitter:

The team working on BiblioBoard has a long history of creating positive disruptive technologies. Chief Business Officer Mitchell Davis, has over twenty years experience in publishing industry and was previously the founder of Amazon CreateSpace (formerly BookSurge) now the largest and most successful indie publishing company in the world. He has been published in dozens of publications writing about the future of publishing and libraries, including: Publishers Weekly, Digital Book World, Insights (JISC), Library Journal, and more. He has presented multiple times at events such as Frankfurt Book Fair, Book Expo American, the American Library Association conferences, Charleston Conference, London Book Fair, DISH (Europeana Annual Conference), and numerous regional conferences.
The software team at BiblioLabs is heavily involved in the open source software community and has been large contributors of code to the Readium open source ePub project.

Problem / Challenge Space:

In 2011, BiblioLabs worked with the British Library to make 80,000 19th Century books from the Library’s collection available on the iPad. These titles had been digitised by Microsoft and represented only about 30% of the collection without comprehensive coverage in any subject area. The associated metadata with the digitised objects was sparse and dated back to the 19th Century. Searching for known items worked but was frequently frustrating because it was impossible to know what titles were included in the project. As a result, it was imperative that we find a way to make the collection easy to browse and engaging to the public. We succeeded through curating the collection to feature high-interest items, putting them in context and displaying them in an elegant multi-media app. For instance, we found all of the titles related to bicycling and organized them into a mini-collection that was featured on the app’s homepage highlighting the many interesting images. This drew the attention of users who most likely did not set out to find old books on bicycling, but they were engaged and shared their excitement with others via social networking sites. As a result, The British Library 19th Century app has been downloaded a quarter of a million times, and it clocks average session lengths close to 20 minutes. These are highly impressive stats in an age in which competition for attention is fierce.
Following this success, other libraries approached BiblioLabs to provide a quote for app development; and we immediately recognized that without tools to curate and publish elegant mobile products at scale, libraries would be unable to afford to promote their digital collections this way. So, we set out to develop a tool set that would make it simple for any subject matter experts (librarians or patrons) to curate multimedia content and make it available on the web on mobile devices in a way that “future proofed” the user-experience as technologies change quickly. Institutions can focus on adding value by content curation, while we keep up with the fast moving mobile technology curves and keep the digital experience ahead of those curves. This seems a perfect public + private partnership. Providing curatorial tools and mobile access that is competitive with the best consumer mobile experiences were central to our endeavor. We were also determined to create a solution that managed all content formats in a single-user interface, rather than forcing patrons into silos that all work differently. Above and beyond these parameters, we knew that the display and ease of use had to rival those of commercial ventures such as Apple, Amazon and Google.

Approach / Methodology:

In designing BiblioBoard’s curatorial tools, our team prioritized making the process simple so no technical expertise would be required to curate beautiful digital exhibits that work on all mobile devices. At the same time, the interface makes it possible to capture extensive metadata. In addition to providing lengthy descriptions, curators can choose item level thumbnails, tag title pages, highlight visually interesting pages, choose from uniform rights statements, give attribution and assign subjects to each item. Content can be uploaded to the system item by item, or it can be integrated via API with existing repositories to provide a federated topic search across digital assets (books, videos, maps, etc.) and further streamline the curatorial process. Further, BiblioBoard works with PDF’s and ePubs as well as most image, audio and video formats. Once content has been selected for an anthology, the curator can organize it into collections, determine the order in which it will appear, enhance the metadata as well as brand the anthology with logos and title if desired.
Another priority for our team was taking the cost out of creating products that are optimized for mobile devices. Since we started by developing for the iPad, we have always been a mobile first company; and with Internet traffic from mobile devices surpassing that from desktops, building for the web alone did not make sense to us. To this end, once someone has used BiblioBoard Creator to curate an anthology, our technology outputs it within native apps for iOS and Android tablet devices. It also creates a web version of the anthology that is fully responsive to screen size giving content consumers multiple choices for mobile access. A large part of making this process sustainable includes the “future-proofing” of the anthologies. With the increasingly rapid release schedules for new operating systems and browsers, it is essential that all share a uniform codebase. So, when Apple for instance updates iOS, the BiblioLabs development team does not need to look at each anthology and revise code separately. Reinventing the wheel for every new digitised collection will not be sustainable at scale for even large research institutions as the wealth of digital materials proliferates and most museums, libraries and archives are increasingly under-resourced. A unified interface that can house disparate content is essential, and it must be updatable without requiring the curators to redo their work or developers to revise code for each individual exhibit.
Most importantly, we wanted the BiblioBoard experience for patrons to compare favorably with other apps they use every day in the consumer world. Cumbersome authentication and checkout procedures would not fly. The interface also needed to allow patrons to interact meaningfully with the content: make lists, bookmark, take notes, export citations, rate and comment, follow and share. With this as our standard, our design team and UI experts have spent countless hours mocking up, testing and enhancing the BiblioBoard interface to ensure users find it both simple and engaging.
The BiblioBoard platform is 100% cloud based with Service Oriented Architecture. It is asynchronous message queue driven, with intelligent auto-scaling algorithms to optimize performance. The mean response time is 800ms in single node benchmark testing for 10,000 concurrent users and 100 concurrent requests. This means we are able to support millions of users. The technology stack reflects our commitment to using the latest, yet proven and stable technologies:
• Languages: Java/JavaScript/Objective-C
• Application Frameworks: Spring/Guice
• Object-Rel Mapping: Hibernate/iOOQ
• Database: MySQL
• App Server: Tomcat
• Search Server: Lucene Solr
• Caches: Redis/Hazelcast/Memcached
We use an Agile development process with sprints/scrums with seamless and frequent deployments. Our process includes continuous code integration, review and QA/testing. To date we have had 470 releases, including 486, 559 lines of code, and have dedicated 113 person-years to BiblioBoard development.

Extent of showcasing BL Digital Content:

The British Library 19th Century books digitised by Microsoft were among the first to be integrated into BiblioBoard Creator, which helped streamline the process of creating British Library anthologies (our term for exhibits) for BiblioBoard. Using this content within BiblioBoard Creator, our editorial team in collaboration with curators from the Library developed a module of 17 anthologies containing 890 eBooks and additional articles, images, and video. Libraries in North America commercially subscribe to these anthologies today, and they are made available in the UK free through the British Library. Following on this success, content curated by the Learning Team at the British Library was added to BiblioBoard as the decision was made to discontinue maintaining the Library’s website where it was being hosted. With the addition of teacher’s guides provided by our editorial team to support the US common core curriculum, BiblioLabs packaged the timeline presentations into The British Library Common Core module. It has since become one of the most popular subscription collections on BiblioBoard in the US; and it is available without charge in the UK. It contains 429 images, 66 documents, 32 audio clips and 13 videos. Today we are exploring how the BiblioBoard infrastructure can be used by the British Library to manage the collection of self-published works from UK authors.

Impact of Project:

Since its release, BiblioBoard has been honored with Publishing Innovation Awards in 2012 and 2013 at the Digital Book World conference and was named Best New Mobile app for libraries by The Charleston Advisor in 2012. It also won a Modern Library Award for design in 2014. It has been featured in articles in Publishers Weekly, Digital Book World, Insights (JISC), and Library Journal. BiblioBoard has also been presented at events such as Frankfurt Book Fair, Book Expo America, the American Library Association conferences, Charleston Conference, London Book Fair, DISH (Europeana Annual Conference), and numerous regional conferences.
BiblioBoard has been selected as the technology infrastructure for statewide digital programs in Massachusetts, North Carolina, Minnesota, Arizona, South Carolina and Kansas. University of Liverpool and University of Sussex are using the technology to curate and host open educational resources. In total, close to 95 million people can access BiblioBoard through their library today, including residents of the UK through the British Library.
Over the last year, British Library content on BiblioBoard has been used over 7,000 times with much of the use in classrooms in the United States. The British Library has earned around $12,000 in royalties to date, and we project sales of British Library content to grow with the rapid expansion of the BiblioBoard customer base.

Issues / Challenges faced during project(s):

Being able to move on when something isn’t working has been a critical element to our success as a company. This has been true for us both on the business side as well as the technology side of the organization. We have frequently had to adjust business models and abandon technology solutions, but we are able to do this without losing site of the end-goal. At some point in the future, born-digital objects may conform to standards and behave predictably, but that is not the case today; and our experience indicates it is unrealistic to expect them to anytime soon. Working in an environment, where we are aggregating digital items from a wide variety of sources and serving them up in a unified interface, we know we will encounter technology hurtles that require creative thinking and often a lot of work to resolve. Early in the development of BiblioBoard, we made the decision to abandon 5 months of work because we realized that HTML5 was not yet ready for prime time when it came to providing a good reading experience. Rather, than live with a less than perfect solution, we started over and built native apps, which offered a far superior result. This experience has shaped our approach whenever we encounter issues. As a rule, we will put in the extra effort to produce a seamless experience for BiblioBoard consumers rather than settle for “good enough.”