Algorithms for mining association rules in bag databases

Ping Yu Hsu, Yen Liang Chen, Chun Ching Ling

Research output: Contribution to journalArticlepeer-review

45 Scopus citations

Abstract

Existing studies in mining association rules in transaction databases assume that a transaction only records the items bought in that particular transaction. However, a typical transaction also records the quantities of items. Because quantity information is not incorporated in the analysis, the association rules cannot reveal what quantities of different items are related with one another. Therefore, this paper reconsiders the conventional transaction database by assuming that each transaction is formed of a set of items as well as their quantities. (We name this extended transaction database as bag database.) In bag databases, algorithms are developed for mining association rules including items' quantities, and three kinds of association rules are generated.

Original languageEnglish
Pages (from-to)31-47
Number of pages17
JournalInformation Sciences
Volume166
Issue number1-4
DOIs
StatePublished - 29 Oct 2004

Keywords

  • Association rule
  • Data mining
  • Fuzzy set

Fingerprint

Dive into the research topics of 'Algorithms for mining association rules in bag databases'. Together they form a unique fingerprint.

Cite this