The American Association for the Advancement of Science (AAAS), publisher of Science, provides an educational resource layer called Science in the Classroom (SitC) that "helps students understand the structure and workings of professional scientific research." It looks like this:

Graduate students provide annotations that are categorized as shown in the Learning Lens widget. Readers select one or more categories to highlight annotated phrases, and click them to read popup annotations.

The workflow supporting this process was cumbersome. To speed up the work and scale it out to a larger team of annotators, AAAS wanted to use Hypothesis. The Hypothesis client was clearly a better way to gather the annotations. But it looks nothing like like Learning Lens. What to do?

Fortunately, the core anchoring libraries used in the Hypothesis client are available as standalone modules (dom-anchor-text-quote, dom-anchor-text-position). That made it possible for the developers of Learning Lens to fetch annotations made using the Hypothesis client and paint them on Science in the Classroom pages in a way that's compatible with Learning Lens. Here's a demo of the technique they used.

And here's an example of the result: Seeing the forest for the trees.

It looks and behaves just like other SitC articles. But the annotations were created with Hypothesis and are dynamically fetched and anchored from Hypothesis when the page loads. It's a nice example of the interoperability that open source software can enable.

There were a few bumps along the way. SitC's editor-in-chief needed to review and approve the annotations created by graduate students. We haven't yet built the group admin capability that will enable that. So we used the Hypothesis API to transfer work from student accounts to the master account from which Learning Lens fetches annotations. We also discovered a problem with overlapping annotations that we solved one way for SitC (and similar use cases) and will address separately in the different context of the Hypothesis client.

All in all, though, we're pleased with the outcome. SitC can now more effectively help students learn how science works. And we've learned valuable lessons about how to put annotation to work. There are a couple of other tools I've been meaning to make available on a prototype basis.