TCS-TR-A-05-9

Date: Tue Sep 27 19:19:30 2005

Title: Finding All Simple Disjoint Decompositions in Frequent Itemset Data

Authors: Shin-ichi Minato

Contact:

  • First name: Shin-ichi
  • Last name: Minato
  • Address: Hokkaido Univ. North 14, West 9, Sapporo 060-0814, Japan
  • Email: minato@ist.hokudai.ac.jp

Abstract. In this paper, we propose a method of finding simple disjoint decompositions in frequent itemset data. The techniques for decomposing Boolean functions have been studied for long time in the area of logic circuit design, and recently, there is a very efficient algorithm to find all possible simple disjoint decompositions for a given Boolean functions based on BDDs (Binary Decision Diagrams). We consider the data model called ``sets of combinations'' instead of Boolean functions, and present a similar efficient algorithm for finding all possible simple disjoint decompositions for a given set of combinations. Our method will be useful for extracting interesting hidden structures from the frequent itemset data on a transaction database. We show some experimental results for conventional benchmark data.


©Copyright 2005 Authors