The ratio of the observed support to that expected if X and Y were independent. Lift > 1 means the two items are positively correlated and the rule is useful.
Measure of rule interestingness.
Used to identify meaningful product associations.