A heuristic rule reduction approach to software fault-proneness prediction

Akito Monden, Jacky Keung, Shuji Morisaki, Yasutaka Kamei, Ken Ichi Matsumoto

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

4 Citations (Scopus)

Abstract

Background: Association rules are more comprehensive and understandable than fault-prone module predictors (such as logistic regression model, random forest and support vector machine). One of the challenges is that there are usually too many similar rules to be extracted by the rule mining. Aim: This paper proposes a rule reduction technique that can eliminate complex (long) and/or similar rules without sacrificing the prediction performance as much as possible. Method: The notion of the method is to removing long and similar rules unless their confidence level as a heuristic is high enough than shorter rules. For example, it starts with selecting rules with shortest length (length=1), and then it continues through the 2nd shortest rules selection (length=2) based on the current confidence level, this process is repeated on the selection for longer rules until no rules are worth included. Result: An empirical experiment has been conducted with the Mylyn and Eclipse PDE datasets. The result of the Mylyn dataset showed the proposed method was able to reduce the number of rules from 1347 down to 13, while the delta of the prediction performance was only. 015 (from. 757 down to. 742) in terms of the F1 prediction criteria. In the experiment with Eclipsed PDE dataset, the proposed method reduced the number of rules from 398 to 12, while the prediction performance even improved (from. 426 to. 441.) Conclusion: The novel technique introduced resolves the rule explosion problem in association rule mining for software proneness prediction, which is significant and provides better understanding of the causes of faulty modules.

Original languageEnglish
Title of host publicationAPSEC 2012 - Proceedings of the 19th Asia-Pacific Software Engineering Conference
PublisherIEEE Computer Society
Pages838-847
Number of pages10
ISBN (Print)9780769549224
DOIs
Publication statusPublished - Jan 1 2012
Event19th Asia-Pacific Software Engineering Conference, APSEC 2012 - Hong Kong, China
Duration: Dec 4 2012Dec 7 2012

Publication series

NameProceedings - Asia-Pacific Software Engineering Conference, APSEC
Volume1
ISSN (Print)1530-1362

Other

Other19th Asia-Pacific Software Engineering Conference, APSEC 2012
CountryChina
CityHong Kong
Period12/4/1212/7/12

Fingerprint

Association rules
Explosions
Support vector machines
Logistics
Experiments

All Science Journal Classification (ASJC) codes

  • Software

Cite this

Monden, A., Keung, J., Morisaki, S., Kamei, Y., & Matsumoto, K. I. (2012). A heuristic rule reduction approach to software fault-proneness prediction. In APSEC 2012 - Proceedings of the 19th Asia-Pacific Software Engineering Conference (pp. 838-847). [6462753] (Proceedings - Asia-Pacific Software Engineering Conference, APSEC; Vol. 1). IEEE Computer Society. https://doi.org/10.1109/APSEC.2012.103

A heuristic rule reduction approach to software fault-proneness prediction. / Monden, Akito; Keung, Jacky; Morisaki, Shuji; Kamei, Yasutaka; Matsumoto, Ken Ichi.

APSEC 2012 - Proceedings of the 19th Asia-Pacific Software Engineering Conference. IEEE Computer Society, 2012. p. 838-847 6462753 (Proceedings - Asia-Pacific Software Engineering Conference, APSEC; Vol. 1).

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

Monden, A, Keung, J, Morisaki, S, Kamei, Y & Matsumoto, KI 2012, A heuristic rule reduction approach to software fault-proneness prediction. in APSEC 2012 - Proceedings of the 19th Asia-Pacific Software Engineering Conference., 6462753, Proceedings - Asia-Pacific Software Engineering Conference, APSEC, vol. 1, IEEE Computer Society, pp. 838-847, 19th Asia-Pacific Software Engineering Conference, APSEC 2012, Hong Kong, China, 12/4/12. https://doi.org/10.1109/APSEC.2012.103
Monden A, Keung J, Morisaki S, Kamei Y, Matsumoto KI. A heuristic rule reduction approach to software fault-proneness prediction. In APSEC 2012 - Proceedings of the 19th Asia-Pacific Software Engineering Conference. IEEE Computer Society. 2012. p. 838-847. 6462753. (Proceedings - Asia-Pacific Software Engineering Conference, APSEC). https://doi.org/10.1109/APSEC.2012.103
Monden, Akito ; Keung, Jacky ; Morisaki, Shuji ; Kamei, Yasutaka ; Matsumoto, Ken Ichi. / A heuristic rule reduction approach to software fault-proneness prediction. APSEC 2012 - Proceedings of the 19th Asia-Pacific Software Engineering Conference. IEEE Computer Society, 2012. pp. 838-847 (Proceedings - Asia-Pacific Software Engineering Conference, APSEC).
@inproceedings{12cb93bb694d41a6854360d023c9dafc,
title = "A heuristic rule reduction approach to software fault-proneness prediction",
abstract = "Background: Association rules are more comprehensive and understandable than fault-prone module predictors (such as logistic regression model, random forest and support vector machine). One of the challenges is that there are usually too many similar rules to be extracted by the rule mining. Aim: This paper proposes a rule reduction technique that can eliminate complex (long) and/or similar rules without sacrificing the prediction performance as much as possible. Method: The notion of the method is to removing long and similar rules unless their confidence level as a heuristic is high enough than shorter rules. For example, it starts with selecting rules with shortest length (length=1), and then it continues through the 2nd shortest rules selection (length=2) based on the current confidence level, this process is repeated on the selection for longer rules until no rules are worth included. Result: An empirical experiment has been conducted with the Mylyn and Eclipse PDE datasets. The result of the Mylyn dataset showed the proposed method was able to reduce the number of rules from 1347 down to 13, while the delta of the prediction performance was only. 015 (from. 757 down to. 742) in terms of the F1 prediction criteria. In the experiment with Eclipsed PDE dataset, the proposed method reduced the number of rules from 398 to 12, while the prediction performance even improved (from. 426 to. 441.) Conclusion: The novel technique introduced resolves the rule explosion problem in association rule mining for software proneness prediction, which is significant and provides better understanding of the causes of faulty modules.",
author = "Akito Monden and Jacky Keung and Shuji Morisaki and Yasutaka Kamei and Matsumoto, {Ken Ichi}",
year = "2012",
month = "1",
day = "1",
doi = "10.1109/APSEC.2012.103",
language = "English",
isbn = "9780769549224",
series = "Proceedings - Asia-Pacific Software Engineering Conference, APSEC",
publisher = "IEEE Computer Society",
pages = "838--847",
booktitle = "APSEC 2012 - Proceedings of the 19th Asia-Pacific Software Engineering Conference",
address = "United States",

}

TY - GEN

T1 - A heuristic rule reduction approach to software fault-proneness prediction

AU - Monden, Akito

AU - Keung, Jacky

AU - Morisaki, Shuji

AU - Kamei, Yasutaka

AU - Matsumoto, Ken Ichi

PY - 2012/1/1

Y1 - 2012/1/1

N2 - Background: Association rules are more comprehensive and understandable than fault-prone module predictors (such as logistic regression model, random forest and support vector machine). One of the challenges is that there are usually too many similar rules to be extracted by the rule mining. Aim: This paper proposes a rule reduction technique that can eliminate complex (long) and/or similar rules without sacrificing the prediction performance as much as possible. Method: The notion of the method is to removing long and similar rules unless their confidence level as a heuristic is high enough than shorter rules. For example, it starts with selecting rules with shortest length (length=1), and then it continues through the 2nd shortest rules selection (length=2) based on the current confidence level, this process is repeated on the selection for longer rules until no rules are worth included. Result: An empirical experiment has been conducted with the Mylyn and Eclipse PDE datasets. The result of the Mylyn dataset showed the proposed method was able to reduce the number of rules from 1347 down to 13, while the delta of the prediction performance was only. 015 (from. 757 down to. 742) in terms of the F1 prediction criteria. In the experiment with Eclipsed PDE dataset, the proposed method reduced the number of rules from 398 to 12, while the prediction performance even improved (from. 426 to. 441.) Conclusion: The novel technique introduced resolves the rule explosion problem in association rule mining for software proneness prediction, which is significant and provides better understanding of the causes of faulty modules.

AB - Background: Association rules are more comprehensive and understandable than fault-prone module predictors (such as logistic regression model, random forest and support vector machine). One of the challenges is that there are usually too many similar rules to be extracted by the rule mining. Aim: This paper proposes a rule reduction technique that can eliminate complex (long) and/or similar rules without sacrificing the prediction performance as much as possible. Method: The notion of the method is to removing long and similar rules unless their confidence level as a heuristic is high enough than shorter rules. For example, it starts with selecting rules with shortest length (length=1), and then it continues through the 2nd shortest rules selection (length=2) based on the current confidence level, this process is repeated on the selection for longer rules until no rules are worth included. Result: An empirical experiment has been conducted with the Mylyn and Eclipse PDE datasets. The result of the Mylyn dataset showed the proposed method was able to reduce the number of rules from 1347 down to 13, while the delta of the prediction performance was only. 015 (from. 757 down to. 742) in terms of the F1 prediction criteria. In the experiment with Eclipsed PDE dataset, the proposed method reduced the number of rules from 398 to 12, while the prediction performance even improved (from. 426 to. 441.) Conclusion: The novel technique introduced resolves the rule explosion problem in association rule mining for software proneness prediction, which is significant and provides better understanding of the causes of faulty modules.

UR - http://www.scopus.com/inward/record.url?scp=84874632768&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84874632768&partnerID=8YFLogxK

U2 - 10.1109/APSEC.2012.103

DO - 10.1109/APSEC.2012.103

M3 - Conference contribution

AN - SCOPUS:84874632768

SN - 9780769549224

T3 - Proceedings - Asia-Pacific Software Engineering Conference, APSEC

SP - 838

EP - 847

BT - APSEC 2012 - Proceedings of the 19th Asia-Pacific Software Engineering Conference

PB - IEEE Computer Society

ER -