Kingsley Idehen's Blog Data Space

If you believe in context so much then the linking open data idea goes bananas :-) why? because "sameAs" is fundamentally wrong.. an entity on dbpedia IS NOT sameAs one on geonames becouse the context is different and bla bla.. so it all crumbles. The hope is that certain entities stay the same no matter, as a unified conceptual "something" regardless of the context. That is assumption of the OKKAM project for example, try to give nmes to things assuming that, for example, San Francisco is San Francisco no matter which context or DB.

but i might have misunderstood the whole thing here ;-)
Giovanni (from a free san francisco wifi picked up in a parking lot ;-))

Giovanni,

Have I given any indication in any of my comments or posts (esp. the ones post Beijing) implying that I condone the manner in which owl:sameAs has been used by DBpedia? What was my whole UMBEL presentation about? My blog posts (including this one) about?

DBpedia was about creating a large corpus of Names via HTTP URIs so as to demonstrate the fundamental utility of Linked Data.

I guess I now need to write a blog post about DBpedia and the role we play in this whole project. Ditto some background history about how the project has evolved.

@Giovanni, I'd like to humbly suggest perhaps you have "misunderstood the whole thing" here, possibly hinging on what is intended as the meaning of "context". Under the intended meaning, context and Linked Data are not in any way at odds, nor are context and sameAs. As used by UMBEL (only one of a myriad of possible contexts, of course, but the one I know best), "context" is provided through a series of reference subject concepts that can be related at the class level with external data and ontologies. The "context" here acts more at how to relate large datasets or data spaces to one another. Subject concepts provide some fixed reference points in a global knowledge space for saying two things "are about" (the predicate is isAbout) the same concept. In other words, a "context" is provided for saying that two disparate resources are asserted to be about the same topic or subject. The isAbout predicate is but another form of RDF link to provide Linked Data "glue". To date, most such Linked Data glue has been provided by sameAs, of course. sameAs appropriately works at the instance level of identical referents. In UMBEL-speak, we call such referent individuals "named entities" and each named entity is related to one or more subject concepts. So, sameAs applies to individuals and assertions about the sameness of two resources referring to the same individual. Individuals, then, are members of subject concept classes, which have graph relationships among them. This scaffolding of instance --> class --> subject concept --> relations between concepts in its entirety provides the reference "context". The statement truly is NOT: "If you believe in context so much then the linking open data idea goes bananas,", but rather, "If you believe in context so much then the linking open data idea gains coherence." Like all language, the meaning of "context" depends on your context (hehe). I think as Kingsley means it, and certainly we do when used for UMBEL, we mean "context" with respect to these reference subject concepts fixed within a global knowledge space, and not different sources describing the same instance resource (e.g., "sameAs") using different characteristics such as your DBpedia and Geonames example.