A new approach to online generation of association rules

We discuss the problem of online mining of association rules in a large dat abase of sales transactions. The online mining is performed by preprocessin g the data effectively in order to make it suitable for repeated online que ries. We store the preprocessed data in such a way that online processing m ay be done by applying a graph theoretic search algorithm whose complexity is proportional to the size of the output. The result is an online algorith m which is independent of the size of the transactional data and the size o f the preprocessed data. The algorithm is almost instantaneous in the size of the output. The algorithm also supports techniques for quickly discoveri ng association rules from large itemsets. The algorithm is capable of findi ng rules with specific items in the antecedent or consequent. These associa tion rules are presented in a compact form, eliminating redundancy. The use of nonredundant association rules helps significantly in the reduction of irrelevant noise in the data mining process.