W3C home > Mailing lists > Public > www-rdf-interest@w3.org > April 2002


From: Danny Ayers <danny666@virgilio.it>
Date: Thu, 18 Apr 2002 12:22:29 +0200
To: <www-rdf-interest@w3.org>
Message-ID: <EBEPLGMHCDOJJJPCFHEFOEEAFLAA.danny666@virgilio.it>

>> The example that came up, was my friend had graphs in his data
>(call graphs
>> from a reverse engineering project). He was trying to get the RDF
>> visualization tools to draw these graphs nicely.


>Dan Connolly has soom tools in this vein, which have produced some nice
>diagrams. It uses TimBL's CWM rdf-logic / rules engine.

The diagram generation is fairly unidirectional :

source data => inference/transformation rules => visualisation

which could cause problems if you want to edit the diagram, and feed the
change back to the source.

The way I've been trying to get around this is to use a general graph model
within the visualisation app, so you have something like :

domain-specific data <=[modelling rules]=> model <=[visualisation rules]=>

with the modelling & visualisation rules specified in RDF (of course), only
rather than using logical implication using an mapping schema that will be
vaguely like XSLT-style transformation, but more able to map both ways - it
gets a lot simpler when you have a generic graph model in the middle.

A bonus with this approach is that the r.h.s. doesn't have to be a
visualisation, it could be any other representation, for instance a
different schema for the same domain information. Transformations/mappings
could be lossy (losing bidirectionality), dependent on the end schema,
though this isn't inherent in the approach.


progress so far :
(since this was last updated I've put together a Swing circles & arrows
front end, though it isn't yet integrated)
Received on Thursday, 18 April 2002 06:42:41 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 22:44:35 UTC