We present our perspective of database mining as the confluence of mac
hine learning techniques and the performance emphasis of database tech
nology. We describe three classes of database mining problems involvin
g classification, associations, and sequences, and argue that these pr
oblems can be uniformly viewed as requiring discovery of rules embedde
d in massive data. We describe a model and some basic operations for t
he process of rule discovery. We show how the database mining problems
we consider map to this model and how they can be solved by using the
basic operations we propose. We give an example of an algorithm for c
lassification obtained by combining the basic rule discovery operation
s. This algorithm not only is efficient in discovering classification
rules but also has accuracy comparable to ID3, one of the current best
classifiers.