Note on:
"An Approach to Discovering Temporal Association Rules",
Juan M. Ale, Gustavo H. Rossi, UNLP, Argentina,
SAC 2000, March 19-21 Como, Italy
This paper addresses the problems of finding interesting rules when
support may be low using temporal information about the transaction
time of the instances in a data set. By lowering support the number of
rules presented to the user increases. Using time as a way to declare
rules obsolete rules are pruned. This does not relate a great deal to
my work.
|