Who writes what checkers? - Learning from bug repositories

Takeshi Yoshimura; Kenji Kono

Who writes what checkers? - Learning from bug repositories

Takeshi Yoshimura, Kenji Kono

Department of Information and Computer Science

Research output: Contribution to conference › Paper › peer-review

2 Citations (Scopus)

Abstract

Static code checkers have been useful for finding bugs in large-scale C code. Domain-specific checkers are particularly effective in finding deep/subtle bugs because they can make use of domain-specific knowledge. To develop domain-specific checkers, however, typical bug patterns in certain domains must first be extracted. This paper explores the use of machine learning to help extract bug patterns from bug repositories. We used natural language processing to analyze over 370,000 bug descriptions of Linux and classified them into 66 clusters. Our preliminary work with this approach is encouraging: by investigating one of the 66 clusters, we were able to identify typical bug patterns in PCI device drivers and developed static code checkers to find them. When applied to the latest version of Linux, the developed checkers found two unknown bugs.

Original language	English
Publication status	Published - 2014
Event	10th Workshop on Hot Topics in System Dependability, HotDep 2014 - Broomfield, United States Duration: 2014 Oct 5 → …

Conference

Conference	10th Workshop on Hot Topics in System Dependability, HotDep 2014
Country/Territory	United States
City	Broomfield
Period	14/10/5 → …

ASJC Scopus subject areas

Hardware and Architecture
Information Systems
Control and Systems Engineering

Cite this

@conference{f28c87d3c9324bbbbedf8a5508f421b4,

title = "Who writes what checkers? - Learning from bug repositories",

abstract = "Static code checkers have been useful for finding bugs in large-scale C code. Domain-specific checkers are particularly effective in finding deep/subtle bugs because they can make use of domain-specific knowledge. To develop domain-specific checkers, however, typical bug patterns in certain domains must first be extracted. This paper explores the use of machine learning to help extract bug patterns from bug repositories. We used natural language processing to analyze over 370,000 bug descriptions of Linux and classified them into 66 clusters. Our preliminary work with this approach is encouraging: by investigating one of the 66 clusters, we were able to identify typical bug patterns in PCI device drivers and developed static code checkers to find them. When applied to the latest version of Linux, the developed checkers found two unknown bugs.",

author = "Takeshi Yoshimura and Kenji Kono",

note = "Funding Information: We thank the anonymous reviewers for their helpful feedback. This work was partially supported by funding from JSPS Research Fellowships for Young Scientists, Hitachi, Ltd. and Nippon Telegraph and Telephone Corportaion (NTT). Publisher Copyright: {\textcopyright} HotDep 2014.; 10th Workshop on Hot Topics in System Dependability, HotDep 2014 ; Conference date: 05-10-2014",

year = "2014",

language = "English",

}

TY - CONF

T1 - Who writes what checkers? - Learning from bug repositories

AU - Yoshimura, Takeshi

AU - Kono, Kenji

N1 - Funding Information: We thank the anonymous reviewers for their helpful feedback. This work was partially supported by funding from JSPS Research Fellowships for Young Scientists, Hitachi, Ltd. and Nippon Telegraph and Telephone Corportaion (NTT). Publisher Copyright: © HotDep 2014.

PY - 2014

Y1 - 2014

N2 - Static code checkers have been useful for finding bugs in large-scale C code. Domain-specific checkers are particularly effective in finding deep/subtle bugs because they can make use of domain-specific knowledge. To develop domain-specific checkers, however, typical bug patterns in certain domains must first be extracted. This paper explores the use of machine learning to help extract bug patterns from bug repositories. We used natural language processing to analyze over 370,000 bug descriptions of Linux and classified them into 66 clusters. Our preliminary work with this approach is encouraging: by investigating one of the 66 clusters, we were able to identify typical bug patterns in PCI device drivers and developed static code checkers to find them. When applied to the latest version of Linux, the developed checkers found two unknown bugs.

AB - Static code checkers have been useful for finding bugs in large-scale C code. Domain-specific checkers are particularly effective in finding deep/subtle bugs because they can make use of domain-specific knowledge. To develop domain-specific checkers, however, typical bug patterns in certain domains must first be extracted. This paper explores the use of machine learning to help extract bug patterns from bug repositories. We used natural language processing to analyze over 370,000 bug descriptions of Linux and classified them into 66 clusters. Our preliminary work with this approach is encouraging: by investigating one of the 66 clusters, we were able to identify typical bug patterns in PCI device drivers and developed static code checkers to find them. When applied to the latest version of Linux, the developed checkers found two unknown bugs.

UR - http://www.scopus.com/inward/record.url?scp=84968837785&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84968837785&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:84968837785

T2 - 10th Workshop on Hot Topics in System Dependability, HotDep 2014

Y2 - 5 October 2014

ER -

Who writes what checkers? - Learning from bug repositories

Abstract

Conference

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this