Using A Tree-Based Method With Transaction Clustering for Mining Association Rules

• Main Authors: Tain-Jung Hsu, 徐添榮
• Other Authors: Yuh-Jiuan Tsay
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/02556452592527872026

碩士 === 國立屏東科技大學 === 資訊管理系 === 93 === Conventional algorithms for mining association rules operate in a combination of smaller large itemsets. This thesis presents a new efficient which combines both the cluster concept and tree structure, while proceeds from mining the maximal large itemsets down to large 2-itemsets, named Tree-based method with transaction Clustering (TC). First, the TC method creates some clusters by reading the database only twice,and then clustering the transaction records to the k-th cluster, where the length of a record is k. The k-th tree is generated based on the k-th cluster only. Then, the large k-itemsets are generated only against all leaves in the tree. Experiments with real-life databases show that TC method outperforms FP-growth method, an efficient and widely used association rule method.