Interestingness measures
An association rule is considered as a strong rule if is satisfies the minimum threshold
criteria, i.e., confidence and support. A minimum support S of a rule A → B indicates
that in x % of all transactions A and B together occurs and it can be calculated using
Eq. (3); whereas a confidence C of a rule indicates that in C % of all transaction when A
occur then B also occurs and it can be calculated using Eq. (4). Lift is another interestingness
measure of a rule, which can be calculated using Eq. (5). A value greater than
1 for the lift measures indicates that the appearance of A and B together is more than
expected whereas a value lower than 1 indicates reverse of the concept. So a rule is considered
as strong if it has a value greater than 1 for the lift parameter.
Support = P(A ∩ B) (3)
Confidence = P(A|B) = (4)