We love what Google has done with it's Blog Search. We love Technorati because it doesn’t just randomly index our blogs, but rather enables blogs and their content to be tagged by people and ranked based on how popular it was (traffic). We simply love the way all the search engines are optimized for searching blogs –one way or another. But amidst all of this, and with the growing popularity of blogging, there is a need that has gone un-answered; The need to find good quality bloggers.
I really can't remember the last time I used Google Blog search or Technorati to find a good Blog. The way I usually find the high quality blogs is through word of mouth, other good blogs, or through blogging communities like Jordan Planet.
Take this scenario; I read a good blogger who reads several other bloggers, so I end up catching one or two of them, who turn out to be really good as well. This is how I was introduced to a large percentage of the blogs I read.
So in addition to all the lightning fast search engines and the neatly tagged directories that are popping up here and there, we need an engine to rank bloggers based on human recommendation.
The base for such an engine would be the OPML (or Blog roll) feed which is currently used for sharing RSS feeds between bloggers. This is because OPML is already being used in a sense for recommending other blogs to readers, and because there is not one good reason why not to use it.
The engine would start with a group of hand-picked blogs which will act as the seeds, or supreme court which will kick-start the recommendation process, resulting in more blogs being added to this (seed group). Seed blogs can recommend other blogs by adding them in their Blog roll, which is picked up by the engine and recorded. Based on these recommendations, any Blog will have the chance to become a seed Blog and play a part in recommending other blogs.
Assuming that the first seed of bloggers are top notch, covering a wide variety of topics, then this engine would hold the largest variety of hand-picked content which is worth reading, anywhere on the web.
I think this would make a nice idea for an open-source project.








