This is an idea that I just had. 

What if you group users in clusters? On Facebook I have friends that are 
relatives, coworkers, other professional contacts, classmates from college, 
dozens of fellow Rush fans, some celebrities with millions of followers, 
and some people that I randomly met. Each of these groups of friends makes 
a cluster which  is possible to identify automatically by analyzing the 
graph of relationships among users. I think, every good book on graphs and 
data structures has a couple of algorithms that does just that.

I believe that an user with 1k relationships may be a member of no more 
than one or two dozens of clusters, thus you could record updates on a 
"Cluster entity and not on the user and when the user queries its updates 
you just have to read the updates on all clusters that the user is a member 
of. 

You could even use a cache for online users, make smaller clusters for most 
active users, organize clusters on an hierarchical mode to make it more 
scalable, and so on. But, of course it is just an idea that I haven't 
tested. 

[]s


-- 
You received this message because you are subscribed to the Google Groups 
"Google App Engine" group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-appengine/-/dRvjC_Oe7MwJ.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/google-appengine?hl=en.

Reply via email to