An efficient way to generate Association Rules with changed minimum support and confidence

Rao, Ashwath B and Herzog, Adrian (2013) An efficient way to generate Association Rules with changed minimum support and confidence. In: Business Intelligence and Knowledge Management (BIAKM-2013), 18th, 19th April, 2013, IBS, Hyderabad .

[img] PDF
Ashwath_Rao_B,_Adrian_Herzog_-_Data_Mining.pdf - Published Version
Restricted to Registered users only

Download (535kB) | Request a copy

Abstract

Mining for association rules is a computation intensive task. This paper proposes a method for speeding up the mining process if association rules are mined on a fixed set of transactions multiple times, while using a different minimum support and/or minimum confidence for each run. The solution is based on a tree data structure which is used to cache previously obtained mining results. The proposed algorithm achieves a remarkable speed gain from the second run onwards. The speedup is especially high when minimum support is increased between two runs. If minimum support is decreased, the speedup is still remarkable.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: Data Mining, Frequent Itemsets, Association Rules, Performance, Cache.
Subjects: Engineering > MIT Manipal > Electronics and Communication
Depositing User: MIT Library
Date Deposited: 30 May 2013 09:36
Last Modified: 30 May 2013 09:36
URI: http://eprints.manipal.edu/id/eprint/136137

Actions (login required)

View Item View Item