Been thinking about what you might call “hyperlocal aggregators” of late. It’s not a novel idea by any stretch but possibly a natural next step. Say you’ve got a load of blogs, forums, etc about a local area. How do you deal with all that stuff? Group blogs like Jon Bounds’ Brumblr network are a possible solution but they do tend to follow the 80/20 rule of participation and still act as a bottleneck. The Hyperlocal Newswire is another possible solution but it’s more search based than perhaps I’m thinking.
I keep returning to I Can Has Cheezbuger for some reason, not so much for the system they use but for the notion of letting stuff bubble up through activity. Or maybe Flickr Interestingness is the thing. Popular isn’t the thing here. I don’t care if lots of people have looked at a thing but I do care if a few people found it really compelling. How do you measure that?
And, ultimately, how can this be connected to a specific geographic place?
Jon Bounds 8:19 pm on August 2, 2009 Permalink
Ah, there are many ways, I hope — I’ve been looking at using the delicious API, and now even the Google Reader ‘like’ system to decide what’s interesting/important in feeds/news/blogs in general. I’ve been wanting to build something that automatically generates the location, as there doesn’t seem to be an easy way for people to describe it in many cases — so I want to generate geo-attention data (places where the information is interesting/useful) in order to feed that back it as a model for place.
I think it’s possible with location brokers (FireEagle/Lattitude) or things like skyhook, to autotag things when people “like” them — we just need people to use the tools that do it (been thinking round a new delicious FF plugin that uses a location service and stores the geo attention data as a tripple tag). Then there’s enough info to work on.
I’m firmly of the opinion that
Jon Bounds 8:30 pm on August 2, 2009 Permalink
… that er, that I was going to delete that bit.