[
Lists Home |
Date Index |
Thread Index
]
- From: uche.ogbuji@fourthought.com
- To: "Bullard, Claude L (Len)" <clbullar@ingr.com>
- Date: Wed, 18 Oct 2000 22:41:24 -0600 (MDT)
I fear I won't have much time for this thread,which is too bad, because it
promises to be a stimulating one.
> Remember, we have OLAP, basic relational techiques,
> etc. and can do quite a bit of reasoning based on
> these. How will the inference engine service improve
> a performance based around these actors?
IMO, OLAP is glorified reporting. There is no synthesis to it. It is a
catalysis that through brute force gives the human some useful tools for
analysis.
But this isn't the key differentiation with RDF. After all, one _could_
superimpose true inference on relational data, yucky as that sounds. This
is, after all what Cognos makes money doing, no?
The main strength of the RDF layer in the closed N-tier system is
extensibility (remember that word that's so often wrongly
capitalized?) RDBMS makes it _very_ difficult to keep the data model
matching the vicissitudes of real-world business. RDF is far more
forgiving and simplifies the task, which admittedly still needs human
intervention in _some_ aspect, of keeping up with changing partnerships,
software, business processes, customer and industry dynamics.
No great magic, but useful stuff if kept in its place.
--
Uche Ogbuji Principal Consultant
uche.ogbuji@fourthought.com +1 303 583 9900 x 101
Fourthought, Inc. http://Fourthought.com
4735 East Walnut St, Ste. C, Boulder, CO 80301-2537, USA
Software-engineering, knowledge-management, XML, CORBA, Linux, Python
|
