Slide 2: Sequential pattern mining: algorithm Slide 4: Sequential pattern mining: Example I (cont.) Home

Sequential Pattern Mining: Example I

---

Given the transaction database with three attributes customer-id, transaction-time, and purchased-items, the mining process were decomposed into the following five phases.

1. Sort Phase The original transaction database is sorted with customer-id as the major key and transaction time as the minor key, the result is set of customer sequences. The table shows the sorted transaction data.

Customer-id Transaction-time Purchased-items 1 1 Oct 23 ’02 Oct 28 ’02 30 90 2 2 2 Oct 18 ’02 Oct 21 ’02 Oct 27 ’02 10, 20 30 40, 60, 70 3 Oct 15 ’02 30, 50, 70 4 4 4 Oct 08 ’02 Oct 16 ’02 Oct 25 ’02 30 40, 70 90 5 Oct 20 ’02 90

2. L-Itemsets Phase
The sorted database is scanned to obtain large (frequent) itemsets according to the predefined support threshold. In order to find the large itemsets, list all itemsets such as{30},{30, 50},{50, 70},{30, 50, 70}, etc. from the above sorted transaction data.

Suppose the minimal support is 40%, in this case the minimal support count is 2, the result of large itemsets is listed in table. For example,

Large Itemsets Mapped to (30) 1 (40) 2 (70) 3 (40, 70) 4 (90) 5

• The itemset {30} is a large itemset because its number of appearance (in Customer IDs 1, 2, 3, and 4) is 4/5 ≥ 40%.


• The itemset {40, 70} is a large itemset because its number of appearance (in Customer IDs 2 and 4) is 2/5 ≥ 40%.


• The itemset {50} is NOT a large itemset because its number of appearance (in Customer ID 3) is 1/5 < 40%.


• The itemset {60, 70} is NOT a large itemset because its number of appearance (in Customer ID 2) is 1/5 < 40%.

Slide 2: Sequential pattern mining: algorithm Slide 4: Sequential pattern mining: Example I (cont.) Home

A Children’s Charity knocked on my door earlier today           asking for a donation to help them build a swimming pool           so I gave them a bucket of water.