Work ongoing at HP Labs in Bristol UK, at their Semantic Web Research Group. where they have done some interesting work that uses the simplicity of blog posting to construct intelligent repositories of what they call 'knowledge snippets' ... information nuggets we would like to store, annnotate and share ... Steve Cayzer of the Labs, in a provocative recent article (link below) discusses the interaction of this simplicty with the ability to store and retrieve knowledge based on embedded intelligence.
Any kind of knowledge management method presupposes you have some critical mass of knowledge to manage. Without having content, no one will use the source. With a simple blog you can easily add and annotate such snippets, but the only way to retrieve is a search. Improvements like Google Desktop Search, make this easier, but it still suffers from the strength of its precision and recall.
Blogs like the one you are reading also contain links, to augment the knowledge they contain by simple reference. This kind of knowledge loading can be improved by adding semantic metadata to the snippits and to their links. The querying and search can be aided by this knowledge. The result is a form of the Semantic Web, overlaid on a simple blog structure.
Of course, this does make the use of the blog structure more complex. It has to be loaded with the semantic knowledge, perhaps through the use of a pre-arranged ontology for classifying and sharing knowledge.
This is important work .. with the potential of linking very simple knowledge storing ideas, like blogs, with more complex knowledge management classification and retrieval methods.
Cayzer also usefully surveys other knowledge-loading mechanisms for blogs. For more detail, and updated research see Cayzer's Semantic Blogging site, which contains a demonstrator for a semantic blog.
And the full December CACM article: Semantic Blogging and Decentralized Knowledge Management, an instructive read. (Full article regrettably only available to ACM subscribers)
Comments