Unified algorithm for undirected discovery of exception rules

Einoshin Suzuki, Jan M. Zytkow

Research output: Contribution to journalArticle

28 Citations (Scopus)

Abstract

This article presents an algorithm that seeks every possible exception rule that violates a commonsense rule and satisfies several assumptions of simplicity. Exception rules, which represent systematic deviation from commonsense rules, are often found interesting. Discovery of pairs that consist of a commonsense rule and an exception rule, resulting from undirected search for unexpected exception rules, was successful in various domains. In the past, however, an exception rule represented a change of conclusion caused by adding an extra condition to the premise of a commonsense rule. That approach formalized only one type of exception and failed to represent other types. To provide a systematic treatment of exceptions, we categorize exception rules into 11 categories, and we propose a unified algorithm for discovering all of them. Preliminary results on 15 real-world datasets provide an empirical proof of effectiveness of our algorithm in discovering interesting knowledge. The empirical results also match our theoretical analysis of exceptions, showing that the 11 types can be partitioned in three classes according to the frequency with which they occur in data.

Original languageEnglish
Pages (from-to)673-691
Number of pages19
JournalInternational Journal of Intelligent Systems
Volume20
Issue number7
DOIs
Publication statusPublished - Jul 1 2005

Fingerprint

Exception
Violate
Simplicity
Theoretical Analysis
Deviation

All Science Journal Classification (ASJC) codes

  • Software
  • Theoretical Computer Science
  • Human-Computer Interaction
  • Artificial Intelligence

Cite this

Unified algorithm for undirected discovery of exception rules. / Suzuki, Einoshin; Zytkow, Jan M.

In: International Journal of Intelligent Systems, Vol. 20, No. 7, 01.07.2005, p. 673-691.

Research output: Contribution to journalArticle

@article{38fdaef4d3af4f308ada966b6cb323ba,
title = "Unified algorithm for undirected discovery of exception rules",
abstract = "This article presents an algorithm that seeks every possible exception rule that violates a commonsense rule and satisfies several assumptions of simplicity. Exception rules, which represent systematic deviation from commonsense rules, are often found interesting. Discovery of pairs that consist of a commonsense rule and an exception rule, resulting from undirected search for unexpected exception rules, was successful in various domains. In the past, however, an exception rule represented a change of conclusion caused by adding an extra condition to the premise of a commonsense rule. That approach formalized only one type of exception and failed to represent other types. To provide a systematic treatment of exceptions, we categorize exception rules into 11 categories, and we propose a unified algorithm for discovering all of them. Preliminary results on 15 real-world datasets provide an empirical proof of effectiveness of our algorithm in discovering interesting knowledge. The empirical results also match our theoretical analysis of exceptions, showing that the 11 types can be partitioned in three classes according to the frequency with which they occur in data.",
author = "Einoshin Suzuki and Zytkow, {Jan M.}",
year = "2005",
month = "7",
day = "1",
doi = "10.1002/int.20090",
language = "English",
volume = "20",
pages = "673--691",
journal = "International Journal of Intelligent Systems",
issn = "0884-8173",
publisher = "John Wiley and Sons Ltd",
number = "7",

}

TY - JOUR

T1 - Unified algorithm for undirected discovery of exception rules

AU - Suzuki, Einoshin

AU - Zytkow, Jan M.

PY - 2005/7/1

Y1 - 2005/7/1

N2 - This article presents an algorithm that seeks every possible exception rule that violates a commonsense rule and satisfies several assumptions of simplicity. Exception rules, which represent systematic deviation from commonsense rules, are often found interesting. Discovery of pairs that consist of a commonsense rule and an exception rule, resulting from undirected search for unexpected exception rules, was successful in various domains. In the past, however, an exception rule represented a change of conclusion caused by adding an extra condition to the premise of a commonsense rule. That approach formalized only one type of exception and failed to represent other types. To provide a systematic treatment of exceptions, we categorize exception rules into 11 categories, and we propose a unified algorithm for discovering all of them. Preliminary results on 15 real-world datasets provide an empirical proof of effectiveness of our algorithm in discovering interesting knowledge. The empirical results also match our theoretical analysis of exceptions, showing that the 11 types can be partitioned in three classes according to the frequency with which they occur in data.

AB - This article presents an algorithm that seeks every possible exception rule that violates a commonsense rule and satisfies several assumptions of simplicity. Exception rules, which represent systematic deviation from commonsense rules, are often found interesting. Discovery of pairs that consist of a commonsense rule and an exception rule, resulting from undirected search for unexpected exception rules, was successful in various domains. In the past, however, an exception rule represented a change of conclusion caused by adding an extra condition to the premise of a commonsense rule. That approach formalized only one type of exception and failed to represent other types. To provide a systematic treatment of exceptions, we categorize exception rules into 11 categories, and we propose a unified algorithm for discovering all of them. Preliminary results on 15 real-world datasets provide an empirical proof of effectiveness of our algorithm in discovering interesting knowledge. The empirical results also match our theoretical analysis of exceptions, showing that the 11 types can be partitioned in three classes according to the frequency with which they occur in data.

UR - http://www.scopus.com/inward/record.url?scp=21844457136&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=21844457136&partnerID=8YFLogxK

U2 - 10.1002/int.20090

DO - 10.1002/int.20090

M3 - Article

AN - SCOPUS:21844457136

VL - 20

SP - 673

EP - 691

JO - International Journal of Intelligent Systems

JF - International Journal of Intelligent Systems

SN - 0884-8173

IS - 7

ER -