Definition: Confilation And Rule

Definition 2.1.4 (Confidence) it adds strength to the implication or rule. If a rule is given by AB then confidence of a rule is given by- Similarly negative association rules are generated. Let A and B be set of items, then negative association rules are generated of the form A ~B, ~A B or ~A ~B. A rule A ~B is valid negative rule if A is frequent itemset and B is an infrequent itemset or And .
Definition 2.1.5 (Support of A ~B) Rule A~B gives the implication as if there A then not possibility of B. So with a support confidence framework, support of A~B is calculated by [5].
Definition 2.1.6 (Confidence of A ~B) It adds strength to a rule A~B .For a rule A~B confidence can be measured by - Where
Definition 2.1.7 (Confidence of ~A B) for a rule ~AB confidence can be measured by -

Definition 2.1.8 (Confidence of ~A ~B) For a rule ~A~B confidence can be measured by - Where If and Support(~A) is given by if .
Once all these negative rules are extracted from the database, their user interestingness is defined on the basis of Piatetsky-Shapiro’s argument [12] of MinInterest.
Definition 2.1.9 (Piatetsky-Shapiro’s argument of Interest) for the set of A and B an interest of B is for A is given by [12]- If then B is positively dependent on A. If then B is negatively dependent on A.
Definition 2.1.10 (Rules of Interest) Uncertainty of association rule is given by one of the measure as Interest. A rule AB is of interest if
Where MinInterest is the minimal interest specified by the user or expert.
Definition 2.1.11 (Valid Positive Rules of Interest)Let for a transactional database D ,set of items are of Interest I and be the itemsets then AB is a valid rule of interest if it satisfies following