‘Canonical knowledge’ it is not a canonical concept so far, or at least there is not enough scholarly production that can be put forward to support it. Therefore, it could not be a DBpedia entity, since it cannot be the topic of a Wikipedia article (which are the source of DBpedia entities). The picture below represents the LOD cloud, that is, a collection of linked and open datasets hosted at Comprehensive Knowledge Archive Network (CKAN). In the very center of the cloud is DBpedia’s dataset.
As an interpretation of this diagram, I will claim that data contained in the cloud are becoming canonical data, just because they occupy the place they occupy, i.e. just because they are hosted at the DataHub from the Open Knowledge Foundation. I will also allege that a correct combination of such canonical data becomes canonical knowledge, as it may be illustrated by this example of data mashup made by SIG.MA. Continue reading