Looks like SharpReader is not the only aggregator that can display relationships between items in RSS feeds. I just found out that Danny Ayers has been working on IdeaGraph, a Java application "for creating visual maps of ideas, that can work with web pages, documents and images".
Among other things, this application can read RSS feeds and show relationships between items. It's unclear from the web-site whether it auto-discovers these relationships or if they need to be added manually, but if Danny has not done this yet, auto-discovery would definitely be a useful feature to add and should be fairly easy to implement. He's already got the display logic in place, and can easily find the relationships by Regex'ing the item descriptions for <a href="..."> tags. And since Java's Regex engine is over twice as fast a .NET's, it seems like a no-brainer to add that logic (if it's not already there).
That reminds me... I need to rerun that benchmark on .NET 1.1 to see if it performs any better than 1.0...
Hi Luke, thanks for the kind words.
The kind of autodiscovery you describe was already on my to-do list, but it just got pushed right up - my plan is for the development of Ideagraph to be driven primarily by feature requests ;-)
I should have a new build ready in the next 3 or 4 weeks, a lot more usable than the current version.
btw, I'm not currently using regex: RSS data is first cleaned (if necessary) to healthy XML, then XSLT'ed to RDF (if necessary, good RSS 1.0 is taken at face value). Relationships like the one you describe are/will be pulled out at that level.
This obviously takes time, but so far it doesn't seem significant compared with the time taken to GET the data. Later on I'll probably add some regex tricks to speed things up, and grab any semantics that might be slipping through the net.
Keep up the good work!
Danny.
Posted by Danny Ayers at April 29, 2003 6:31 AMWelcome the Idea Applications
Ideagraph .