2019 ACM/IEEE International Symposium on Empirical Software Engineering and Measurement (ESEM) 2019
DOI: 10.1109/esem.2019.8870178
|View full text |Cite
|
Sign up to set email alerts
|

Revisiting and Improving SZZ Implementations

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
14
0

Year Published

2020
2020
2022
2022

Publication Types

Select...
3
2
2

Relationship

0
7

Authors

Journals

citations
Cited by 29 publications
(14 citation statements)
references
References 39 publications
0
14
0
Order By: Relevance
“…We are not aware of an SZZ variant that takes this into account. Additionally, we ignore changes that can be explained by refactorings similar to the work by Neto et al (2018Neto et al ( , 2019. Based on the results by Mills et al ( 2018), these modifications should be able to account for about 94.6% of the false positive bug inducing changes.…”
Section: Improving Defect Labelingmentioning
confidence: 99%
See 2 more Smart Citations
“…We are not aware of an SZZ variant that takes this into account. Additionally, we ignore changes that can be explained by refactorings similar to the work by Neto et al (2018Neto et al ( , 2019. Based on the results by Mills et al ( 2018), these modifications should be able to account for about 94.6% of the false positive bug inducing changes.…”
Section: Improving Defect Labelingmentioning
confidence: 99%
“…Moreover, we described how the inducing change can be used for assigning defects to releases in Section 4.1. To this aim, we compare four approaches for the identification of bug inducing changes: 1) the standard SZZ algorithm 2) JLMIV, i.e., our improved linking with the issue validation, but standard SZZ to determine inducing changes; 3) JLMIV+R, i.e., the improvement to ignore changes to non-java files, whitespace only changes, documentation changes, and refactorings; 4) JLMIV+AV that further extends JLMIV by using the affected versions field; and 5) SZZ-RA, i.e., a state of the art variant of SZZ based on the work by Neto et al (2018Neto et al ( , 2019. Our variant of SZZ-RA differs from the original work only in details: we added a filter that ignores test and documentation files, and use the links between commits and issues based on the Jira issue pattern.…”
Section: Bug Inducing Changesmentioning
confidence: 99%
See 1 more Smart Citation
“…Bug-Inducing com-Percentage of submitted commits that mits (SZZ)*,|1 introduce bugs [43]. * 2 i Pull-request (GitHub) or change-request (Gerrit) how the percentage of contributors who declared themselves as female increased to 33% compared to the 18% (20%) values for Cat-1 (Cat-2).…”
Section: Ed Ia N _ # a Ttem P Ts E F T -# Actual _c Om M Itsmentioning
confidence: 99%
“…Another threat to construct validity is related to the accuracy of the SZZ heuristic used to identify fault-inducing commits. Although this heuristic does not achieve a 100% accuracy, it has been successfully employed and reported to achieve good results in multiple empirical studies from the literature [60][61][62]. We also did a manual validation of the bug inducing commits as described in Section 3.3.2 by inspecting the changes of a small sample of bug inducing commits.…”
Section: Threats To Validitymentioning
confidence: 99%