Scheduled discovery of exception rules

Research output: Chapter in Book/Report/Conference proceedingConference contribution

15 Citations (Scopus)

Abstract

This paper presents an algorithm for discovering pairs of an exception rule and a common sense rule under a prespecified schedule. An exception rule, which represents a regularity of exceptions to a common sense rule, often exhibits interestingness. Discovery of pairs of an exception rule and a common sense rule has been successful in various domains. In this method, however, both the number of discovered rules and time needed for discovery depend on the values of thresholds, and an appropriate choice of them requires expertise on the data set and on the discovery algorithm. In order to circumvent this problem, we propose two scheduling policies for updating values of these thresholds based on a novel data structure. The data structure consists of multiple balanced search-trees, and efficiently manages discovered patterns with multiple indices. One of the policies represents a full specification of updating by the user, and the other iteratively improves a threshold value by deleting the worst pattern with respect to its corresponding index. Preliminary results on four real-world data sets are highly promising. Our algorithm settled values of thresholds appropriately, and discovered interesting exception-rules from all these data sets.

Original languageEnglish
Title of host publicationDiscovery Science - 2nd International Conference, DS 1999, Proceedings
EditorsSetsuo Arikawa, Koichi Furukawa
PublisherSpringer Verlag
Pages184-195
Number of pages12
ISBN (Print)354066713X, 9783540667131
DOIs
Publication statusPublished - Jan 1 1999
Externally publishedYes
Event2nd International Conference on Discovery Science, DS 1999 - Tokyo, Japan
Duration: Dec 6 1999Dec 8 1999

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1721
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other2nd International Conference on Discovery Science, DS 1999
CountryJapan
CityTokyo
Period12/6/9912/8/99

Fingerprint

Exception
Data structures
Scheduling
Specifications
Updating
Data Structures
Scheduling Policy
Search Trees
Expertise
Threshold Value
Schedule
Regularity
Specification

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Suzuki, E. (1999). Scheduled discovery of exception rules. In S. Arikawa, & K. Furukawa (Eds.), Discovery Science - 2nd International Conference, DS 1999, Proceedings (pp. 184-195). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1721). Springer Verlag. https://doi.org/10.1007/3-540-46846-3_17

Scheduled discovery of exception rules. / Suzuki, Einoshin.

Discovery Science - 2nd International Conference, DS 1999, Proceedings. ed. / Setsuo Arikawa; Koichi Furukawa. Springer Verlag, 1999. p. 184-195 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1721).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Suzuki, E 1999, Scheduled discovery of exception rules. in S Arikawa & K Furukawa (eds), Discovery Science - 2nd International Conference, DS 1999, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1721, Springer Verlag, pp. 184-195, 2nd International Conference on Discovery Science, DS 1999, Tokyo, Japan, 12/6/99. https://doi.org/10.1007/3-540-46846-3_17
Suzuki E. Scheduled discovery of exception rules. In Arikawa S, Furukawa K, editors, Discovery Science - 2nd International Conference, DS 1999, Proceedings. Springer Verlag. 1999. p. 184-195. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/3-540-46846-3_17
Suzuki, Einoshin. / Scheduled discovery of exception rules. Discovery Science - 2nd International Conference, DS 1999, Proceedings. editor / Setsuo Arikawa ; Koichi Furukawa. Springer Verlag, 1999. pp. 184-195 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{95251fbbdc8940e1afd956b284fb325c,
title = "Scheduled discovery of exception rules",
abstract = "This paper presents an algorithm for discovering pairs of an exception rule and a common sense rule under a prespecified schedule. An exception rule, which represents a regularity of exceptions to a common sense rule, often exhibits interestingness. Discovery of pairs of an exception rule and a common sense rule has been successful in various domains. In this method, however, both the number of discovered rules and time needed for discovery depend on the values of thresholds, and an appropriate choice of them requires expertise on the data set and on the discovery algorithm. In order to circumvent this problem, we propose two scheduling policies for updating values of these thresholds based on a novel data structure. The data structure consists of multiple balanced search-trees, and efficiently manages discovered patterns with multiple indices. One of the policies represents a full specification of updating by the user, and the other iteratively improves a threshold value by deleting the worst pattern with respect to its corresponding index. Preliminary results on four real-world data sets are highly promising. Our algorithm settled values of thresholds appropriately, and discovered interesting exception-rules from all these data sets.",
author = "Einoshin Suzuki",
year = "1999",
month = "1",
day = "1",
doi = "10.1007/3-540-46846-3_17",
language = "English",
isbn = "354066713X",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "184--195",
editor = "Setsuo Arikawa and Koichi Furukawa",
booktitle = "Discovery Science - 2nd International Conference, DS 1999, Proceedings",
address = "Germany",

}

TY - GEN

T1 - Scheduled discovery of exception rules

AU - Suzuki, Einoshin

PY - 1999/1/1

Y1 - 1999/1/1

N2 - This paper presents an algorithm for discovering pairs of an exception rule and a common sense rule under a prespecified schedule. An exception rule, which represents a regularity of exceptions to a common sense rule, often exhibits interestingness. Discovery of pairs of an exception rule and a common sense rule has been successful in various domains. In this method, however, both the number of discovered rules and time needed for discovery depend on the values of thresholds, and an appropriate choice of them requires expertise on the data set and on the discovery algorithm. In order to circumvent this problem, we propose two scheduling policies for updating values of these thresholds based on a novel data structure. The data structure consists of multiple balanced search-trees, and efficiently manages discovered patterns with multiple indices. One of the policies represents a full specification of updating by the user, and the other iteratively improves a threshold value by deleting the worst pattern with respect to its corresponding index. Preliminary results on four real-world data sets are highly promising. Our algorithm settled values of thresholds appropriately, and discovered interesting exception-rules from all these data sets.

AB - This paper presents an algorithm for discovering pairs of an exception rule and a common sense rule under a prespecified schedule. An exception rule, which represents a regularity of exceptions to a common sense rule, often exhibits interestingness. Discovery of pairs of an exception rule and a common sense rule has been successful in various domains. In this method, however, both the number of discovered rules and time needed for discovery depend on the values of thresholds, and an appropriate choice of them requires expertise on the data set and on the discovery algorithm. In order to circumvent this problem, we propose two scheduling policies for updating values of these thresholds based on a novel data structure. The data structure consists of multiple balanced search-trees, and efficiently manages discovered patterns with multiple indices. One of the policies represents a full specification of updating by the user, and the other iteratively improves a threshold value by deleting the worst pattern with respect to its corresponding index. Preliminary results on four real-world data sets are highly promising. Our algorithm settled values of thresholds appropriately, and discovered interesting exception-rules from all these data sets.

UR - http://www.scopus.com/inward/record.url?scp=84957811911&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84957811911&partnerID=8YFLogxK

U2 - 10.1007/3-540-46846-3_17

DO - 10.1007/3-540-46846-3_17

M3 - Conference contribution

SN - 354066713X

SN - 9783540667131

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 184

EP - 195

BT - Discovery Science - 2nd International Conference, DS 1999, Proceedings

A2 - Arikawa, Setsuo

A2 - Furukawa, Koichi

PB - Springer Verlag

ER -