The Mining Algorithm of Maximum Frequent Itemsets Based on Frequent Pattern Tree

Table 1

Pseudocode for constructing new FP-tree.

---

Algorithm 1: The construction algorithm of new FP-tree.

---

Input: Transaction database D, minimum support min_sup. In the process of constructing the FP-tree, the node-pre of each node is null, and the node-pre is assigned after the FP-tree is constructed.

Output: FP-tree

Process:

Scan database D to get frequent 1-itemsets, sort them from large to small according to the support degree, and get the frequent item sequence FI₁;

Create a new FP-tree node, assign the value of the node_name to null, and use it as the root node T;

Create FP-tree: call Create function

Function Create()

For each transaction in the database T_i do

{Initialize the current pointer in the FP-tree to point to the root node T;

Put the items in the transaction T_i that meet the minimum support count requirement into the queue Q in descending order of support count;

For each item in the queue Q do

FP-tree. Insert(I); //Insert I in the child of the node pointed to by the current pointer (if it exists, its counter will be incremented by 1) and move the current pointer to the node

}

Function Insert (char name)//Insert a node with an item named name into the FP-tree

{if the node pointed to by the current pointer has no child nodes or there is no node whose node_name field is name in the child nodes

Then

{Create a new FP-tree node whose node_name field is name；

current- > node_children = node;

current = = node;//Modify the current pointer to point to the new node

current- > node_link = item_head;

item_head = current;//Add it to the item_name linked list whose item_name value is name in the header table

}

else increase the node_count value of the child node by 1 and modify the current pointer to point to it;

}

---