System and method for mining generalized association rules in databases

摘要: A system and method for discovering consumer purchasing tendencies includes a computer-implemented program which identifies transaction itemsets that are stored in database appear the user-defined minimum number of times, referred to as support. The contain items characterized by hierarchical taxonomy. Then, discovers association rules, potentially across different levels taxonomy, comparing times each large appears particular subsets itemset database. When relationship exceeds predetermined confidence value, outputs generalized rule is representative consumers. set rules can be pruned uninteresting i.e., do not occur at frequency significantly than what expected based upon occurrence rule's ancestors.