|
Title:
|
MINING AND FILTERING HIDDEN ASSOCIATION RULES |
|
Author(s):
|
Marco-Antonio Balderas Cepeda, Martín-Eleno Vogel Vázquez |
|
ISBN:
|
978-972-8939-67-0 |
|
Editors:
|
Piet Kommers and Pedro Isaías |
|
Year:
|
2012 |
|
Edition:
|
Single |
|
Keywords:
|
Association, patterns, rules, anomalous, filter, databases |
|
Type:
|
Full Paper |
|
First Page:
|
417 |
|
Last Page:
|
423 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
Traditional association rules are useful to discover potentially interesting patterns in databases, but the discovery of infrequent/rare patterns is difficult (if not impossible) because in most cases infrequent patterns are hidden from the traditional definition of association rule (AR). The anomalous association rules (anomalous rules) are association rules representing a rare and uncommon behavior that deviates from a frequent common pattern. In this paper, we develop filtering techniques, to discover compact and significant anomalous rule-sets. Moreover, in addition to reduce the number of anomalous rules obtained, the approach proposed can be parallelized immediately. Furthermore, the rules obtained can be found useful in Web Mining, social networks mining, bio-surveillance, credit screening, and any application that requires the identification of rare and significant patterns. This approach substantially reduces the number of rules obtained in about 97%. In addition to reducing the obtained rules, the approach can be applied to knowledge discovery of full-correlated rare patterns. |
|
|
|
|
|
|
