I’m not an expert on ontologies, the semantic web, metadata or controlled vocabularies, but I’ve had enough conversations with enough experts to know that more control will not address our information management needs. Recent conversations with people smarter than me have me concluding that Smart Search is (will be) an excellent tool and that the RDF standard seems to be quite useful with its minimalist approach. From the CETIS Report (MS Word Doc, page 23):
I recently asked if metadata was dead and received some good advice:
- From Anol: "Problems with folders and metadata – that’s a closed system, somebody else define the taxonomy. Theory of entropy proves itself when the closed system of folders and metadata goes into a complete chaotic mode."
- and from Keith, "Maybe metadata structures are dying, but there’s a distinct difference between metadata and metadata structures. If you’re going to ask, "Is metadata dead?" why not also ask, "Is tagging [with METADATA!] dead?"
After perusing the 121 pages of the two CETIS reports [I didn’t read every item], I came away with the feeling that trying to control chaos is a losing game. Instead of asking how many angels can dance on the head of a pin, it’s time for the learning industry as a whole to shift its effort to more pragmatic solutions, because the answers from CETIS et al are not very clear. Having watched the enormous efforts ($$$) that the military, academia and corporations have put into metadata and controlled learning structures, without any measurable improvements in learning or performance outcomes, I have to ask if this is worth the time and money. My suggestions:
- like Lego, use the simplest of basic structures (RDF?)
- build better search into online learning applications
- only build taxonomies, ontologies & controlled vocabularies based on a specific user need, not "just-in-case"
- give learners and facilitators more tools to manage their information (tags, tagclouds, smart search, etc)
- focus on tools to surf the chaos, not control it