I’m very happy (though feeling a bit outplayed) to find a great example of a micro material out in the vast internet. It’s a bit similar to an idea that I had (see previous post), though to be fair, it is much more nicely designed than anything I’ve ever done.
It’s an online service that automatically creates an annotated version of the text, a word list, and a vocabulary quiz from basically any website you put into it. It’s very cool and basically automates something that students (and teachers) traditionally have done by hand.
Given that it’s creating the materials on a per-request basis (rather than preprocessing a bunch of webpages and having the materials available on demand), it’s not surprising that it takes about 20-30 seconds to have them ready. Still, that’s a very impressive automation pipeline to have set up!
I’m not exactly sure which word lists the app is checking to identify “difficult” vocabulary, though presumably it’s a frequency-based list like the AWL (Coxhead, 2000), or perhaps something more recent.
I had initially thought that these definitions were being constructed from the text itself (ie, using the word context and semantic inference), though it looks like it’s just querying the Oxford Dictionary API to get definitions, then putting those into both a word list and a vocab quiz (still an impressive feat!).
I’m also not exactly sure how the application is getting around the Terms of Service for a lot of the web sites it gets content from. A lot of places expressly forbid reproducing their content on other sites, and that’s what the re-formatted article is, essentially.
At any rate, a very cool example of something highly focused that can be (and has been) successfully automated. My hat’s off to the folks at word-booster, and I hope to have something similar (though based on comprehension questions rather than pure vocabulary definitions) at some point in the future.