How difficult it would be to implement multiple bucket
classifers for the same message (aka gmail labels - but
done automatically)?
I would like to help with this taks; unless somebody tells
me it is too complex to do based on what the software
does today.
Logged In: YES
user_id=578491
It wouldn't be hard since we have the data about all the
bucket classifications for a message but the core of POPFile
is around trying to differentiate between buckets. I
suppose we could say add the top 3 classifications to a
message, but I'm not sure about this.
Logged In: YES
user_id=262888
I am talking about "bucketizing" a message on more than one
dimension. (say priority 1 vs priority 2 vs priority 3 on one
hand and work vs personal on the other hand)
this is different than just assigning top 3 buckets to a
message. essentially, we look at each bucket and verify if the
message passes a certain threshhold for it. if it does we
assign it but we always test for all the buckets.
does this make sense?