CHANGE/CLARIFICATIOn to QN II:--Collaborative filtering

[Subbarao Kambhampati <rao>]

For the question on Collaborative Filtering (Qn II),
 I had said that
    you can use significance weight to be 1 uniformly.

 INSTEAD of that 

 You should use significance weighting to be min (1 , n/3) where n is the 
number of items co-rated by the current user and the active user.


(The answers make more sense with this better significance weighting... )

Rao


ps: Thanks to Ryan Wilson for bringing this to my notice.