This article feels like a 'guest post' promoting a specific product, though it allows me to address some wider points. There's this space where graphs and concept maps interact and as I've been working on an open graph for gRSShopper I've come across a number of approaches, including not only concept maps but also more restrictive approaches like ontologies and linked data. As I read this article (or, more accurately, looked at the images) I was struck by the vagueness of the lables being used in the connectors - things as varied as 'the most important is' to 'is not just a' to 'means'. My feeling right now is that this approach is too loose - you're not going to find any useful associations if you just use lables willy-nilly. By the same token, I find Semantic Web style lables too restrictive. Not every association is a semantic association. I think, overall, that there are different types of connectors for different purposes (though this may one of those claims that is either false or trivial).
Today: 1 Total: 1089 [Direct link] [Share]
