How to Execute the Research Paper by Anita de Waard.
I had to create the category, “dynamic updating,” to at least partially capture what Anita describes in this presentation. I would have loved to be present to see it in person!
The gist of the presentation is that we need to create mechanisms to support research papers being dynamically linked to the literature and other resources. One example that Anita uses is linking a patient’s medical records to reports in literature with professional tools for the diagnostician.
It isn’t clear how Linked Data (no matter how generously described by Jeni Tennison) could be the second technology for making research papers linked to other data. In part because as Jeni points out, URIs are simply more names for some subject. We don’t know if that name is for the resource or something the resource represents. Makes reliable linking rather difficult.
BTW, the web lost its ability to grow in a “gradual and sustainable way” when RDF/Linked Data introduced the notion that URIs cannot be allowed to fail. If you try to reason based on something that fails, the reasoner falls on its side. Not nearly as robust as allowing semantic 404′s.
Anita’s third step, an integrated workflow is certainly the goal to which we should be striving. I am less convinced about the mechanisms, such as generating linked data stores in addition to the documents we already have, are the way forward. For documents, for instance, why do we need to repeat data they already possess? Why can’t documents represent their contents themselves? Oh, because that isn’t how Linked Data/RDF stores work.
Still, I would highly recommend this slide deck and that you catch any presentation by Anita that you can.