In any event, I believe the design of our "viewer engine" has to have some sort of
intelligence for searching by using ontologies as the techical backbone for the
various tasks of information integration and mediation. This is explained by
important conclusions drawn from the following research projects and papers:
* Tim Berners-Lee's Semantic Web Road map. This document gives a road map - a
sequence for the incremental introduction of technology to take us, step by
step, from the Web of today to a Web in which machine reasoning will be
ubiquitous and devastatingly powerful.
http://www.w3.org/DesignIssues/Semantic.html
* OntoWeb - Ontology-based information exchange for knowledge management and
electronic commerce http://www.cs.vu.nl/~ying/OntoWeb
* On-To-Knowledge - recent state-of-the-art reports: Information Extraction &
Representation and Query Languages for Semistructured
Data
http://www.ontoknowledge.org/countd/countdown.cgi?OTK-EnerSearch-10May2000.ppt
http://www.ontoknowledge.org/downl/del5.pdf
http://www.ontoknowledge.org/countd/countdown.cgi?draftdel8.pdf
Conclusions from the "Representation and Query Languages for Semistructured Data"
paper says:
"In chapter 1 we concluded that RDF was to be preferred as a representation language
over XML. The main reason for this is that RDF explicitly commits to specific
ontological modelling primitives, whereas XML only provides syntactic constructions
without any ontological commitments. The choice for RDF also aligns the
On-To-Knowledge project with the W3C initiatives related to
the Semantic Web, and with the choices made in the recently started DARPA funded
research programme DAML. This choice for RDF as the data-representation language
immediately rules out any of the XML-based query-languages (since we required that
the query-language should closely fit to the underlying data-model). Of the RDF
query-languages known to us at the time of writing, RQL seems by far the most
promising candidate, and fulfills most if not all of our requirements. One proviso
must be made with this choice: RQL firmly commits to the RDF/RDF-Schema data model.
This is of course as required, but also implies a limitation: any extensions to the
RDF data-model (as is the intended use of RDF-Schema) cannot be directly queried in
RQL. Consider for example the case of OIL: it extends the basic RDF/RDF-Schema model
with new primitives (e.g. cardinality constraints). Such new elements in the
data-model can only be included in a syntactic way in RQL queries, but the RQL query
language is not able to process the semantics of such new primitives. Three
alternatives exist as a way out of this dilemma:
* define a language which allows for some kind of ifsemantic plug-in
definitionsls, in order include the semantics of such schema-extensions into
the query language
* implement a new or extended query-language for each extension of the data-model
* write a query-compiler wich compiles queries for the extended language into a
set of queries for the original language.
Our choice for RQL as the query-language in the On-To-Knowledge architecture implies
that we recommend the third of these options."
Eugene Eric Kim wrote:
> On Mon, 7 Aug 2000, John J. Deneen wrote:
>
> > So in regards to the "viewing engine", SHOE is a small extension to HTML which
> > allows web page authors to annotate their web documents with machine-readable
> > knowledge. SHOE makes real intelligent agent software on the web possible.
>
> Just to keep things clear, the "viewing engine" is a human viewing
> engine. As has been mentioned before, metacontent is an important aspect
> of the system, and SHOE may very well be an integral part of how we do
> that. Additionally, metacontent could certainly affect different
> views. And we certainly want to make sure that machines can view our
> data, just as humans, although I suspect that we can just send machines
> raw XML rather than put it through any transformation. However, for the
> immediate purposes of designing the first version of the viewing engine, I
> don't think SHOE holds much relevance.
>
> -Eugene
>
> --
> +=== Eugene Eric Kim ===== eekim@eekim.com ===== http://www.eekim.com/ ===+
> | "Writer's block is a fancy term made up by whiners so they |
> +===== can have an excuse to drink alcohol." --Steve Martin ===========+
This archive was generated by hypermail 2.0.0 : Tue Aug 21 2001 - 17:57:49 PDT