websights

Putting Pages to Pasture: Rethinking Learning Content

For hundreds of years, we’ve been pouring educational content onto sheets of paper. It’s remarkable to think that virtually everything we know, every human idea, creation, and discovery, has been stamped out into 2D text and diagrams. And with all that practice, we’ve gotten pretty good at it. We’ve also grown accustomed to consuming information this way. And although it’s rarely the best way to do it, studying the March on Washington or the works of Beethoven from the confines of a book feels normal. You might say we’ve been book trained.

So it shouldn’t be a surprise that we’ve clung to the familiarity and safety of the page even as we’ve tiptoed into the era of ultraportable digital readers. Is this the best we can do? Today, we have an amazing opportunity to redefine not only the notion of a page, but the very notion of learning content itself.

If you had the chance to go back in time to invent the modern cell phone, would you give each phone an archaic 10 digit number? Or would you use something semantically relevant and easy to remember, like an e-mail address? Would you keep the numeric keypad or would you design it with a keyboard? Or even voice control? You’d be awfully unlikely, knowing what you know now, to design it the way it was originally designed.

Moving beyond the page has put virtually every imaginable medium at our fingertips: audio, video, 3D objects, interactive quizzes, and many things we’ve not yet thought up. It has also allowed us to move away from the traditional, linear list of pages into a graph of content objects that can be explored by breadth or depth. All the objects in the graph are indexed and ranked in order to replace a traditional page-numbered index, with a search engine that allows the learner to more quickly and efficiently find what they’re looking for. (For those who still need page numbers, Inkling even has a map from page numbers in the print version to locations in the digital content, so getting around the old fashioned way is even easier than with the original book.) With the freedom to organize content in any way, we’re limited only by what the underlying platform can support.

When we first began designing Inkling, we evaluated open specifications for representing, describing and packaging e-books. Most suffered from the same constraints as their paper counterparts, and weren’t really designed with “unbook” content in mind. So we instead developed our own content and packaging language, and we designed it from the ground up for object-oriented content. We named it S9ML (after Inkling’s birth name, Standard Nine).

As its name suggests, S9ML is an XML specification. It covers the core components of every title that we produce, such as the object graph, the glossary and the search index. It also describes an interface for each content object type–from multi-figure sequences to interactive assessments. S9ML is simply a markup language that we use to semantically describe the content in our titles. It doesn’t prescribe the details of how content is presented, and it’s agnostic of the underlying platform. That’s important for two reasons. First, Inkling content is portable, which means it can be displayed on an iPad today and an Android tablet or iPhone tomorrow. Second, it means that the content will fit the screen resolution, controls, and user interface guidelines of each platform we target, effectively guaranteeing a first-class user experience on any device Inkling supports.

What’s more, because the S9ML for a given title includes the search index, indexing and ranking is done before the content arrives on the device. That means that Inkling’s search is fast and relevance-based, rather than being slow and linear like most other reading platform search functions. It also means that search results are guaranteed to be the same on all devices that run Inkling.

Inkling makes content available by the chapter, unit, or the entire book. Under the hood, that means we’re segmenting the graph of content into smaller disjoint sub-graphs that can be downloaded individually. In reality, the sub-graphs aren’t actually constrained to chapters or unit boundaries. It’s simply a reasonable way to parcel out the content for users. And as books become a less important output of the publishing process, we’ll have even more flexibility in how we package and distribute content, as small as a single node or as large as the whole content graph.

As the specification matures, we’ll begin sharing S9ML with the world.

Of course, S9ML and the many technical challenges we solve each day at Inkling are only the beginning. We continue to think about what’s needed to build, package, distribute, and display the learning content of the future. And we’re excited to tackle the engaging and important technical challenges we’ll face as we improve the world of digital content for everybody who loves to learn.