Useful list, almost a roadmap or set of scenarios. The last two might be especially intriguing, in view of the Semantic Web.

Ah, yes! It may sound technical to some, but there’s something very useful about deep linking which can help fulfill Berners-Lee’s Semantic Web idea much more appropriately than what is currently available. Despite so many advances in Web publishing (and the growing interest in Linked Open Data), it’s often difficult to link directly to an online item of interest. In a way, Hypothesis almost allows readers to add anchor tags to an element so it can be used in a direct link.

Well-structured annotations can pave the way towards Linked Open Data.

L’interopérabilité est essentielle à l’avènement du Web des données liées (en éducation comme ailleurs).

The “Open World Assumption” is central to this post and to the actual shift in paradigm when it comes to moving from documents to data. People/institutions have an alleged interest in protecting the way their assets are described. Even libraries. The Open World Assumption makes it sound quite chaotic, to some ears. And claims that machine learning will solve everything tend not to help the unconvinced too much. Something to note is that this ability to say something about a third party’s resource connects really well with Web annotations (which do more than “add metadata” to those resources) and with the fact that no-cost access to some item of content isn’t the end of the openness.

C’était long, en 2002! Où en est la BnF, aujourd’hui? D’une certaine façon, ce résumé semble prédire la venue des données, la fédération des catalogues, etc. Pourtant, il semble demeurer de nombreux obstacles, malgré tout ce temps. Et si on pouvait annoter le Web directement?

Post, comment, annotation… All different, but can all have the same predicate.

Standards matter.

Which is where OER catalogues (tied to the Semantic Web) may shine. Sure, they can require a lot of work. But this is precisely why they matter.

As long as it has some URI, it can be annotated. Any object in the world can be described through the Semantic Web. Especially with Linked Open Data.

Waiting for those to be promoted, through Hypothesis and other Open Annotations platforms.

Maps make a great case for SVG. There are some neat libraries and tools to play with SVG maps but, more importantly, maps make it easy to understand that an image can be semantic.

A couple of weeks before Shepazu posted this, was playing with SVG maps of contemporary Africa’s political boundaries. (Especially those used on Wikipedia; including some which separate South Sudan.) Been teaching African Studies (on occasion) for years, and maps of the continent tend to become important quite quickly.

Those SVG maps with which I started playing were pretty neat in several respects. The fact that they were vector drawings instead of bitmaps meant that they easily be resized without causing visual artifacts. More importantly, though, each country was drawn as a named outline, so it was possible to play with them as separate objects.

One thing I was trying to do is create an animation which would show where each country fits in a region of the continent, using this United Nations geoscheme. Doing so, eventually noticed that Sudan and South Sudan had been classified as part of different regions, which is an interesting tidbit which could lead to useful classroom discussions.

Haven’t retraced all the steps but, at some point, I’ve used a Public Domain map of Africa from Wikimedia Commons (itself based on another Public Domain map), and ended up creating a simple animated version using Tumult’s Hype commercial HTML5 editor.

It’s flawed in many ways, but for someone with almost no background in this things, it’s a significant accomplishment.

(Surely, the same could be done through SVG itself. Haven’t been able to learn how to do so.)

Playing with those maps taught me quite a few things. For instance, the benefits of a well-tagged image. And some rudimentary notions of CSS-based animations. Or the limitations linked to selecting rectangular sections of an image (with a large overlap between Northern and Western Africa, for instance).

The experience also gave me all sorts of ideas. Such as annotating parts of a well-structured image. Or uses for Open Street Maps. Or ways to embed interactive content (including quizzes) in Open Textbooks.

The key point, perhaps, and what led me to Schepers’s work (including this deeply insightful SVG-based presentation and interactive infographic about annotations) is that Open Standards can open up fascinating opportunities for learning.

So nice to be working at a standards-happy learning technology non-profit!

Ominous…

Be sure not to leave a trailing space on this URL

I think that this type of approach will be necessary for tagging methods.

Yes, I believe that this is the best compromise.

Read the whole essay it is worth the while...