Discovering Time Weighted Association Rules

시간 가중치를 고려한 연관규칙

  • 손승현 (한양대학교 산업공학과) ;
  • 김재련 (한양대학교 산업공학과)
  • Published : 2000.12.01


Discovery of association rules has been used useful in many fields, especially in the fields of the inventory display, catalog design and cross selling. In previous works, all transactions In the database are treated uniformly. In this paper, we present a method for partitioning transactions in the database using time weights. Transactions are assigned different weights as time goes on. Examples show that these method provides purchasing patterns in the database as well as finding association rules.