Discovering Significant Sequential Patterns in Data Stream by an Efficient Two-Phase Procedure

Algorithm 2

GreateTree.

	Input: DS: dataset
	: support threshold
	xs: testable support
	Output: a tree T and a header table H
(1)	Initiate a header table H containing the fields of the item, support, and links
(2)	For each transaction Td of DS, do
(3)	For each item X in Td do
(4)	Calculate H.X.support
(5)	End For
(6)	End For
(7)	Delete unpromising items from H with support and xs constraints
(8)	Initialize a Tree T with an empty root node
(9)	For each transaction Td of DS, do
(10)	Delete unpromising items from Td
(11)	Insert the sequential itemset S of Td
(12)	For each item X in S
(13)	Update H.X.support
(14)	Add the links
(15)	End For
(16)	End For
(17)	Return T and H