- From: John Flynn <jflynn@bbn.com>
- Date: Tue, 5 Apr 2005 14:04:33 -0400
- To: <semantic-web@w3.org>
- Cc: <jflynn@bbn.com>
- Message-Id: <200504051805.j35I5NvF023490@aragorn.bbn.com>
The following is a concept for a very simple (poor man's) tagging scheme for marking up entities (instances) imbedded within html web pages. The terms that are to be designated as instances of classes or properties of some external rdfs/owl ontology are marked up using an html-like tag <eN uri of ontology:class/property-name /eN>, where N is the number of words in the instance name. This simple entity tag provides the information needed to locate the relevant ontology and the class or property within that ontology to which the instance applies. It also prevents the entity markup from displaying on the rendered web page. Using entity tags would prevent the need to embed rdfs/owl in the html page but still allow for the simple association of instances with ontologies in situations where that might be useful. Spiders could be designed to collect and catalog the resulting ontology and instance information for use by query engines and/or agents. This represents only very early thinking about such a concept, but with some refinement it might provide a simple way for people who are only familiar with html web site creation to begin to become involved in the development of the Semantic Web. Additional description and some examples of entity markup using the FOAF and Wine ontologies, and an example of the cataloged results, are at (1). (1) http://semanticsimulations.com/EML.htm John Flynn
Received on Tuesday, 5 April 2005 18:05:41 UTC