The Creative Commons and the Association of Educational Publishers (AEP) announced a new metadata initiative today, aiming to provide a framework for tagging and organizing educational content online -- a framework that will work with the newly proposed Schema.org semantic markup.
Metadata. The Semantic Web. I can sense yours eyes glazing over, reader. But bear with me. This is a very big deal, and although it sounds on the surface like (geeky) good news -- it comes with the promise of making (open) educational information easier to find -- I have a few concerns.
Let me explain.
The Politics of Markup
Earlier this week, three of the world's major search engines -- Google, Bing, and Yahoo -- announced Schema.org, a "shared vocabulary" with which website owners can mark up their content. With over 100 types of new tags, webmasters will be able to add semantic data to their websites and in turn, search engines will have a better sense of what these pages contain.
At first blush, this seems like a win for the semantic Web. But ReadWriteWeb's editor Richard MacManus wrote a very interesting post this morning, asking some important questions about the new project and asking if, in fact, it's simply a "Google land grab.
His concerns stem from the fact that the new markup proposed as part of Schema.org is a big-3-search-engine effort, something that will compete with existing open standards. Rather than thinking about how semantic markup can help structure the Web, we may be seeing how search engines' needs will structure the Web.
Furthermore, while open standards use RDFa (a version of the semantic markup sponsored by the World Wide Web Consortium), Schema.org has opted to go with Microdata, Google's own spec.
Marking Up Educational Content
So what does this have to do with better searching for educational information?
The announcement today from Creative Commons and AEP marks the first industry-specific implementation of Schema. It's aimed at making educational content more easily discoverable.
That's an important move, because while we often say that you can learn anything you want from the Internet, it can actually be quite difficult -- when equipped with just a search engine -- to find good, quality, level-appropriate educational content. It's challenging to separate what's a good scholarly resource from what's a commercial resource, or what's an age-appropriate resource, or what's an openly-licensed resource.
So hooray for semantic markup! Right?
What's Educational versus What's Optimized for SEO
From the press release that accompanied today's announcement:
"The initiative to create a metadata framework for learning resources has already gained support from several leading commercial and non-commercial organizations. These include BetterLesson, Curriki, Houghton Mifflin Harcourt, the Institute for the Study of Knowledge Management in Education (ISKME), McGraw-Hill Education, Monterey Institute for Technology in Education (MITE), Pearson, Promethean, Scholastic Inc., and SMART Technologies Inc. The launch partners have agreed to contribute to the development of the specifications and to support their implementation."
In other words, all the big educational content providers are on board. That bodes well for widespread adoption, again a win for thorough markup. But what about the "little" providers -- um, teachers? Well, get ready for a new semantic-based SEO battle, I guess, folks. And that's the cautionary note I leave you with, in the midst of what is otherwise some exciting news.
Don't get me wrong: I'm pleased to see people on board with better markup -- because I love the idea of the semantic web and I think we have hardly begun to make the connections necessary in order to make the most of metadata.
But I'm not yet convinced that an industry-led effort or a search-engine led effort is the way to do this. I can't help but worry that we'll see educational content echo something that happened when Google introduced its rich snippets markup to recipes. Shortly after Google unveiled the recipe search, food blogger Amanda Hesser observed that "better recipe search" had little to do with "better food." She noted that the top result for cassoulet came from the mega-site Epicurious -- not surprising. Furthermore, by using the filters Google provided for recipe searching, you could look for cassoulets that could be made in 15 minutes (um, no) or one that would only be 77 calories (WTF). Hesser said, "I fear to contemplate the future of American cooking."
How can we make sure that semantic markup of educational content doesn't take rich, complex and traditional content -- let cassoulet be a metaphor here -- and reduce it to the point of inanity and tastelessness?