PLANMINE: Predicting plan failures using sequence mining

Mohammed J. Zaki, Neal Lesh, Mitsunori Ogihara

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

This paper presents the PLANMINE sequence mining algorithm to extract patterns of events that predict failures in databases of plan executions. New techniques were needed because previous data mining algorithms were overwhelmed by the staggering number of very frequent, but entirely unpredictive patterns that exist in the plan database. This paper combines several techniques for pruning out unpredictive and redundant patterns which reduce the size of the returned rule set by more than three orders of magnitude. PLANMINE has also been fully integrated into two real-world planning systems. We experimentally evaluate the rules discovered by PLANMINE, and show that they are extremely useful for understanding and improving plans, as well as for building monitors that raise alarms before failures happen.

Original languageEnglish
Pages (from-to)421-446
Number of pages26
JournalArtificial Intelligence Review
Volume14
Issue number6
DOIs
Publication statusPublished - 1 Dec 2000
Externally publishedYes

Fingerprint

Data mining
Planning
planning
event
Data Base
Data Mining
Monitor
Real World

Keywords

  • Plan monitoring
  • Predicting failures
  • Sequence mining

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Artificial Intelligence

Cite this

PLANMINE : Predicting plan failures using sequence mining. / Zaki, Mohammed J.; Lesh, Neal; Ogihara, Mitsunori.

In: Artificial Intelligence Review, Vol. 14, No. 6, 01.12.2000, p. 421-446.

Research output: Contribution to journalArticle

Zaki, Mohammed J. ; Lesh, Neal ; Ogihara, Mitsunori. / PLANMINE : Predicting plan failures using sequence mining. In: Artificial Intelligence Review. 2000 ; Vol. 14, No. 6. pp. 421-446.
@article{caa0f3efc8854d5fbaeeb36882b5b29f,
title = "PLANMINE: Predicting plan failures using sequence mining",
abstract = "This paper presents the PLANMINE sequence mining algorithm to extract patterns of events that predict failures in databases of plan executions. New techniques were needed because previous data mining algorithms were overwhelmed by the staggering number of very frequent, but entirely unpredictive patterns that exist in the plan database. This paper combines several techniques for pruning out unpredictive and redundant patterns which reduce the size of the returned rule set by more than three orders of magnitude. PLANMINE has also been fully integrated into two real-world planning systems. We experimentally evaluate the rules discovered by PLANMINE, and show that they are extremely useful for understanding and improving plans, as well as for building monitors that raise alarms before failures happen.",
keywords = "Plan monitoring, Predicting failures, Sequence mining",
author = "Zaki, {Mohammed J.} and Neal Lesh and Mitsunori Ogihara",
year = "2000",
month = "12",
day = "1",
doi = "10.1023/A:1006612804250",
language = "English",
volume = "14",
pages = "421--446",
journal = "Artificial Intelligence Review",
issn = "0269-2821",
publisher = "Springer Netherlands",
number = "6",

}

TY - JOUR

T1 - PLANMINE

T2 - Predicting plan failures using sequence mining

AU - Zaki, Mohammed J.

AU - Lesh, Neal

AU - Ogihara, Mitsunori

PY - 2000/12/1

Y1 - 2000/12/1

N2 - This paper presents the PLANMINE sequence mining algorithm to extract patterns of events that predict failures in databases of plan executions. New techniques were needed because previous data mining algorithms were overwhelmed by the staggering number of very frequent, but entirely unpredictive patterns that exist in the plan database. This paper combines several techniques for pruning out unpredictive and redundant patterns which reduce the size of the returned rule set by more than three orders of magnitude. PLANMINE has also been fully integrated into two real-world planning systems. We experimentally evaluate the rules discovered by PLANMINE, and show that they are extremely useful for understanding and improving plans, as well as for building monitors that raise alarms before failures happen.

AB - This paper presents the PLANMINE sequence mining algorithm to extract patterns of events that predict failures in databases of plan executions. New techniques were needed because previous data mining algorithms were overwhelmed by the staggering number of very frequent, but entirely unpredictive patterns that exist in the plan database. This paper combines several techniques for pruning out unpredictive and redundant patterns which reduce the size of the returned rule set by more than three orders of magnitude. PLANMINE has also been fully integrated into two real-world planning systems. We experimentally evaluate the rules discovered by PLANMINE, and show that they are extremely useful for understanding and improving plans, as well as for building monitors that raise alarms before failures happen.

KW - Plan monitoring

KW - Predicting failures

KW - Sequence mining

UR - http://www.scopus.com/inward/record.url?scp=0034457239&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0034457239&partnerID=8YFLogxK

U2 - 10.1023/A:1006612804250

DO - 10.1023/A:1006612804250

M3 - Article

AN - SCOPUS:0034457239

VL - 14

SP - 421

EP - 446

JO - Artificial Intelligence Review

JF - Artificial Intelligence Review

SN - 0269-2821

IS - 6

ER -