PubSub LinkRank
I was just in the processing of setting up some PubSub feeds for internal purposes when I saw their new LinkRanks feature. Wow. Very, very interesting (you can also read the about complete with mathematical formulas.
What does it do? It tracks the popularity of links to domains (domains, not individual pages). It specifically only looks at links that are within the body of syndicated content (which is pretty much limited to weblogs at this point). So, all those linkrolls and sidebars don't count.
Also, if you include a link like this:
http://psi.pubsub.com/20040413:linkranks:1
Then PubSub will track the "conversation" around their LinkRanks topic. This is another feature which doesn't look to be "live" yet, but seems very, very interesting.
The potential here is to have true communication between and around web pages and blogs -- by pointing to a common link, all posts around a particular topic could be gathered together in one place.
Similar
Comments
PubSub LinkRanks
Anonymous wrote:
This problem is much more serious when you index all links from an HTML blog rather than just the links from within the items or entries in the blog's RSS or Atom file. PubSub *only* indexes links that appear in actual entries... While bloggers tend to put links to tool makers in their "sidebars", they link less frequently to the tool makers when they actually write blog entries.
If you can think of any way for us to improve the PubSub LinkRanks, please feel free to send a mail with your suggestions to: feedback AT pubsub.com
bobwyman
CTO, PubSub Concepts, Inc. http://bobwyman.pubsub.com
link ranks
i think one thing which i have noticed is that a lot of top ranking sites in any such blog indices are the tool makers and as a result you really don't get the right information. i am assuming that most of the people who are not listed in the pubsub database don't get counted in any such ranking. i might be wrong.
PubSub currently parses only
PubSub currently parses only sites with feeds, if that is what you mean. I believe they actually look for information from all the major "ping" servers -- e.g. weblogs.com etc. etc. -- and generally try to be as inclusive as possible.
Remember -- these are not blog indices. These count the outbound links, not the inbound links. So if a bunch of people link to a New York Times story (which is not a blog and does not have an RSS feed), then that will move "up the charts". And in fact, the NYT is third on the list, and there are many other news organizations listed.
It has been stated before that the blog world (which pretty much matches up with those sites that have feeds available, although this is changing) is a bit inward looking, so these results might be skewed in that sense.
I believe that Technorati actually removes some of the "tool makers" (e.g. blog software providers like MT, Wordpress, etc.) from their indices. I would rather have it all in and do my own analyses.