Mining Sequential Patterns

We are given a large database of customer transac tions where each transaction consists of customer id transaction time and the items bought in the transac tion We introduce the problem of mining sequential patterns over such databases We present three algo rithms to solve this problem and empirically evalu ate their performance using synthetic data Two of the proposed algorithms AprioriSome and Apriori All have comparable performance albeit AprioriSome performs a little better when the minimum number of customers that must support a sequential pattern is low Scale up experiments show that both Apri oriSome and AprioriAll scale linearly with the num ber of customer transactions They also have excel lent scale up properties with respect to the number of transactions per customer and the number of items in a transaction