We did some 'auto query generation' in our SemanticSever. The user draws
a graph with some nodes labeled with a question mark, then the engine
first builds the query for the users to make sure this is what they
intended, then allows them to run the query. S-Server searches on many
participating physical servers from the list you can maintain - it
treats the Web as one "information space" hosted on many servers.
You can try S-Server directly at our site - www.semanticsoft.net. After
login as a guest, the first thing you see is a "desktop" with a Start
button like in Windows, but you are actually on our minuature
SemanticWeb hosted on our servers. Press Start button and select Help -
this will help you to use this tool.
The "canvas" (or "tissue", "media", "substratum") on which our Semantic
Web is built, allows ontologies to be stored in a "transparent" manner
so that the query engines "see inside" them while regular digital
artifacts are kept as blobs. "Folders" are just a UI simulation of
Windows on this "canvas" - behind the UI you actually group resources
(things) and, this manner, change the "topology" of the information space.
I would treat "understanding" of some content (like the content in GGG)
as representation of that content onto such "canvas", because as soon as
the content is onto the "canvas", the engines can query inside such
content and they are also aware of the context which is also on the
canvas as Semantic Web formatted content.
Ioachim Drugus, Ph.D.
Main Software Architect,
SemanticSoft, Inc.
www.semanticsoft.net
रविंदर ठाकुर (ravinder thakur) wrote:
i think one of the important feature would be the query processor for
automatic query formation. this will involve making query processor
UNDERSTAND the data in GGG and the context in which query is asked.
however in the semantic web i havn't found anything directed at
automatic query formation. extracting any peice of information
requires knowledge of structure of data(ontology) and hand coding the
query to SQARQL equivalant. has any work being been done previously on
auto query generators ?
On Wed, Dec 17, 2008 at 4:07 AM, Juan Sequeda <[email protected]
<mailto:[email protected]>> wrote:
IMO, the (object->fact->links) is what RDF/RDFS/OWL does. So that
complements those layers of the cake.
The objective of the cake (the semantic web) is to allow
serendipity and discovery. So I retract and what I said before.
Discovery cannot be a layer; it should be Inference. The fifth
layer is trust. The whole cake will allow discovery and serendipity
Juan Sequeda, Ph.D Student
Dept. of Computer Sciences
The University of Texas at Austin
www.juansequeda.com <http://www.juansequeda.com>
www.semanticwebaustin.org <http://www.semanticwebaustin.org>
On Tue, Dec 16, 2008 at 1:02 PM, Knud Hinnerk Möller
<[email protected] <mailto:[email protected]>> wrote:
On 16.12.2008, at 17:27, Kingsley Idehen wrote:
Knud et al,
I think Ravinder has started the process of fixing the
current Semantic Web layer cake :-) Which is a very good
thing (imho, but not seeking a Layer Cake discussion
explosion).
I'd rather say his proposal (objects->facts->links->...) is
complementary to the current SW layer cake. It shows what is
going on conceptually, whereas the current SW cake (which I
agree probably needs to be fixed) is more of a technology
stack. An interesting paper I read related to this is:
A. Gerber, A. van der Merwe, and A. Barnard. A functional
Semantic Web architecture. In Proceedings of the 5th European
Semantic Web Conference (ESWC2008), Tenerife, Spain, pages
273–287. Springer, June 2008.
The tricky part is the interchangeable nature of
"Discovery" and "Trust" in any such scheme layer-wise. For
instance, do "Discovery" and "Trust" occupy Layers 4, 5 or
either ? We ultimately want to reason against trusted data
sources, but the serendipity quotient of discovery is a
key factor re. the dynamic nature of "trusted sources".
Since I am clearly thinking and writing (aloud) at the
same time, I would suggest:
Layer 4 - Discovery (with high Serendipity Quotient)
Layer 5 - Trust (albeit inherently volatile)
Kingsley
I'm not sure discovery and trust belong in this stack at all.
Not that I don't think they are extremely important, but what
I see in Ravinder's stack is a description of the nature of
data on the SW. Of course, what I see might not be what he
intended! :) The next layer should describe how the data is
different from the previous layers. As I pointed out in a
previous mail, I think this difference could be in inferenced
data vs. explicit data.
Cheers,
Knud
-------------------------------------------------
Knud Möller, MA
+353 - 91 - 495086
Smile Group: http://smile.deri.ie
Digital Enterprise Research Institute
National University of Ireland, Galway
Institiúid Taighde na Fiontraíochta Digití
Ollscoil na hÉireann, Gaillimh
