Hey,
I need an algorithm similar to amazon.com which would give me n items
that were rated highest by the users that rated item x high. i.e. people
who liked this book also liked those 5..
So,
N = number of items.
X = number of users.
For a given items i and user u, R(i,u) | {Rmin <= R(i,u) <= Rmax, R(i,u)
belongs to N+} is the rating of the item i by the user u.
What i need is:
Given an item i, what are the m (m=5 for example) items that were ranked
highest by users that ranked i high?
There are some ambiguities in this quite informal problem description:
1. What does it mean item ranked high by a user? imho it means R(i,u) >
avg ( R(i,u) for all u, i), i.e. higher than the average of all ratings
by all users.
2. What does it mean items that were ranked highest by users that ranked
item i high? No idea. Help me interpret that please.
Thanks!
P.S.
If you reply, please reply to me as well, not just the newsgroup.
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
