Data MiningAssociation Rule

1
Data Mining-Association Rule

• CS 157B, Spring Semester 2007
• By
• Nathanael Chow

2
Overview

• What is data mining
• Association Rule
• Support
• Confidence
• Example
• Important of association rule
• References

3
What is Data Mining?

• Data mining is the process of searching large
  volumes of data for patterns, correlations, and
  trends.

4
Associations

• Frequently co-occurring features
• Example someone who buys bread is quiet likely
  also to buy milk
• Amazon customer who buys Gladiator will also
  buys Brave Heart

5
Association (cont.)
6
Association Rule
7
Association Rule Example
8
Association Rule Example (cont.)
9
Association Rule Example (Cont.)

• Pizza ? Soda min_support 50
• min_conf 50
• Support pizza U soda 2/4 50
• Confidence (pizza U soda) / (pizza) 2/4
• 50
• Conclude Pizza ? Soda

10
Association Rule Example (Cont.)

• Soda ? Pizza min_support 50
• min_conf 50
• Support soda U pizza 2 / 4 50
• Confidence (soda U pizza) / (soda) 2 / 2
• 100
• Conclude Soda ? Pizza

11
Association Rule Example (Cont.)
12
Important of knowing association rule

• Supper market to get a better profit(sale) by
  placing the associate items next to each other
• Online sale higher sales by placing real time
  advertising

13
(No Transcript)
14
References

• A. Silberschatz, H. Korth, S. Sudarshan, 2006.
  Database System Concepts Fifth Edition
• http//cs.sjsu.edu/7Elee/cs157b/cs157b.html
• http//www.cs.sjsu.edu/faculty/lee/cs157/cs157a.ht
  ml