What pruning property helps generate high-confidence rules from one frequent itemset?

A. If an antecedent is frequent, every rule is valid B. If a consequent fails minimum confidence, its supersets can be pruned in standard rule generation C. Confidence rises whenever the consequent grows D. All consequents have equal confidence