[FoRK] Q re: ConceptNet (also FluidDB)
sdw at lig.net
Wed Oct 21 12:24:25 PDT 2009
Jeff Bone wrote:
> Re: JAR's comments...
> Yup to all, particularly the graph theoretical stuff. Important to
> note (of course you realize this already, just noting "for the record)
> that rdf / semantic networks, your usual induction and inference
> engines, and so forth are all just specialized graph engines with some
> strict constraints on the meanings of nodes and edges and the
> algorithms which operate on them. IMHO the innovation in ConceptNet
> is in having a higher-order / fuzzier semantics for the nodes and
> edges. CN's concepts and relations operate at a level a lot closer to
> the fuzzy semantic / analogical reasoning of humans (and hence the
> real uses of natural language) than e.g. the stricter systems.
I view a lot of the semantic technology as A) a better way to organize
information for straight business purposes and B) often the best form of
bottom-up feeder information that could feed into knowledge and
probabilistic graph reasoning systems. When you can't or don't need to
capture precise probabilistic knowledge, RDF et al is just fine and far
simpler. Currently however, most tools are annoyingly formal and
stiff. We need projects like ConceptNet to make it easy to capture and
use knowledge, hopefully spurring multiple hot applications that get a
good feedback loop going.
> But at some level it's all graph theory, whether we're talking CN, rdf
> / cwm / SparQL / etc., Prolog-like systems, and so on.
> Neural networks and other similar systems are something else entirely,
> though, and while there's a mapping here it's a bit elusive.
> Spreading activation in semantic networks with fuzzy, defeasible
> semantics seems like a pretty rich topic at present.
I'm glad you now see a mapping / equivalence. I would really like to
focus on proving what my intuition tells me, but have several more
pressing concerns first. Someone will eventually map the equivalence of
NN with probabilistic relational graph reasoning (Bayesian / Markov with
imprecise reasoning, etc.) and both will be A) unified or B) both be
stronger as a result.
More information about the FoRK