Tuesday, February 21, 2006
Generative Methodology Glass Bead Games
On the limits of the OWL standard à [184]
Reading material [1]
Reading material [2]
Reading material [3]
Summary of the discussion up to this point à [186]
Dialogue about ontology mapping
Statement: The selling point made by RDBMS advocates is
that RDBMS will provide
persistence and efficient engines for very large Semantic Web linked OWL
ontologies.
Response: I'm a little bit confused here. Does it mean that these efficient engines do the work much in the manner that reasoners like Racer, Fact++, Pellet etc do? A database is still a database but not a knowledge base or an ontology. so what ever query tools that are built for databases, they still try to answer the queries with the existing closed world like reasoning. So, my question is, those efficient engines, that the RDBMS advocates are claiming to provide, do they do reasoning(draw inferences) of the open world assumption style?
Statement: A file that size is likely to cause problems
in the in-memory (file
based) projects in Protege.
This is really what the database backend has been designed to handle.
We have ongoing experiments for streaming a file directly into the
protege database model - bypassing the application - for really large files.
The db can then be hooked up to a ProtegeOWL database project.
Response: I
don't understand some thing here. If owl based ontologies are to be streamed
into database models, then how do we use the inference engines for reasoning
the ontology? Do the reasoners like Racer can also deal with protege database
models(converted owl ontologies)??
Statement: I'll ask what the current state of this is
(a few people here have much
more experience with pushing Protege's memory boundaries than me).
Also, what is a reasonable upper limit on ontology size.
Response: I'm also interested to know the upper limit of the size of an owl ontology.