Monday, July 05, 2004
New approaches to social networks
Google has a couple of tricks up it's sleeve - link: and related:
Now, I'm sure you all know about these - link shows which pages link to a particular page, and related shows pages that are 'similar' (whatever that actually means). However, what should happen is that we combine these implicit social links with the explicit ones we already have (links on the website, people in the blogroll, perhaps even those sites in the favourites), assess them to see how many refer to each other, cluster them and then display them as a proper 'neighbourhood'. It's a neat idea, huh? Bits have been done - there's a Python script out there that does the related parts - but all really need to be in there, and a decent representation used, and then it could be a very nice way of illustrating what and where I look, and who looks at me - it maps out my interests and so on. If anyone writes it in php or Java, please let me know.....
Google has a couple of tricks up it's sleeve - link: and related:
Now, I'm sure you all know about these - link shows which pages link to a particular page, and related shows pages that are 'similar' (whatever that actually means). However, what should happen is that we combine these implicit social links with the explicit ones we already have (links on the website, people in the blogroll, perhaps even those sites in the favourites), assess them to see how many refer to each other, cluster them and then display them as a proper 'neighbourhood'. It's a neat idea, huh? Bits have been done - there's a Python script out there that does the related parts - but all really need to be in there, and a decent representation used, and then it could be a very nice way of illustrating what and where I look, and who looks at me - it maps out my interests and so on. If anyone writes it in php or Java, please let me know.....
Comments:
Post a Comment
(c) 2003-2005 Russell Beale
Atom
RSS