[jdev] APML and PubSub

[Jonathan Dickinson]

Hi All,

I thought someone might want to do something interesting with this technology.

""APML allows users to share their own personal Attention Profile in much the same way that OPML allows the exchange of reading lists between News Readers. The idea is to compress all forms of Attention Data into a portable file format containing a description of ranked user interests.""

Given an APML binding for each PubSub node (e.g. "user interface") a PubSub service could provide a user with a list of suggested nodes based on an APML document that they provide (via IQ or something): or automatically send them publications in any node based on each publications' tags.

Furthermore, if the PubSub spec got a little 'smarter' the author could set the 'affinity' for each post's tag: allowing them to indicate how well the post relates to the particular tag. This could be used to filter publications based on the "value" attribute in the APML. Thus:

Known APMLs:
Fred:
 Food 0.6
 User Interface 0.9
Jane
 Food 0.8
 User Interface 0.3

Subscriber Creates Publication:
 Food (3/5 = 0.6 - Invert = 0.4)
 Sends to  Fred and Jane
Another one:
 User Interface (2/5 = 0.4 - Invert = 0.6)
 Sends to  Fred

Just some hashing out of some ideas; I don't have the time to really go at it. Just wanted to throw it out there and see what you all think: maybe someone will hop on it and write a XEP.

<http://www.apml.org/>

-- Jonathan

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.jabber.org/jdev/attachments/20090401/b6ae5a4e/attachment-0003.htm>