- From: Jozef De Roo <jozef.deroo@skynet.be>
- Date: Thu, 2 Dec 1999 23:58:29 +0100
- To: <www-rdf-interest@w3.org>
- Message-ID: <000a01bf3d18$c08ae220$4f09eec3@skynet.be>
During the past year we have thougt a lot about RDF (and the semantic Web). We also did some experimental work. RDF can be a lot of fun! The idea was to generate an RDF proof for a question about a given set of RDF facts and rules. The basic concepts are: 1. A fact is a simple RDF assertion like the property p of subject s is object o. 2. A rule is also an RDF assertion but the subject is a predicate and the object is a kind of lambda expression. 3. This set of RDF information is acquired from the Web and translated into Euler nodes to form a kind of logic program. 5. The deduction engine only follows Euler paths (the concept Euler found several hundred years ago) so that endless deductions are avoided. This means that no special attention have to be paid to recursions or to graph merging. 6. A kind of access limited logic is used to limit the combinatorial explosion during deduction. More information can be found at ftp://windsor.agfa.be/outgoing/RCEI/NET/euler/index.html or ftp://windsor.agfa.be/outgoing/RCEI/NET/euler.zip This is a very experimental thing and we are looking for ... flaws. jos.deroo.jd@belgium.agfa.com --- Agfa Research Centre on Electronic Imaging
Received on Thursday, 2 December 1999 17:59:56 UTC