Interesting idea. I suspect a very simple LSA would be sufficient to note blog entries that are similar or are repeats of the same info, and to cluster them into topics (a la Google News). Likewise some knowledge that people often include links to where info came from ("seen on Slashdot") could be used to enhance results. Talk to me some more about it if you want...I did some of this text analysis stuff on a just-ended project, so some of it is fresh in my brain.
But then I think I might want a hierarchy, or a generalized graph, which I navigate to find new information. I already think this way; I know Boingboing aggregates from, say, Gizmodo, which pulls from Engadget, so as you do I only skim Engadget if I've read the other two.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
(no subject)
Date: 2007-07-31 03:12 pm (UTC)But then I think I might want a hierarchy, or a generalized graph, which I navigate to find new information. I already think this way; I know Boingboing aggregates from, say, Gizmodo, which pulls from Engadget, so as you do I only skim Engadget if I've read the other two.