Proceedings of the 19th International Conference on Evaluation and Assessment in Software Engineering 2015
DOI: 10.1145/2745802.2745833
|View full text |Cite
|
Sign up to set email alerts
|

Towards an automation of the traceability of bugs from development logs

Abstract: Context: Information and tracking of defects can be severely incomplete in almost every Open Source project, resulting in a reduced traceability of defects into the development logs (i.e., version control commit logs). In particular, defect data often appears not in sync when considering what developers logged as their actions. Synchronizing or completing the missing data of the bug repositories, with the logs detailing the actions of developers, would benefit various branches of empirical software engineering… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

1
7
0

Year Published

2018
2018
2023
2023

Publication Types

Select...
2
1

Relationship

1
2

Authors

Journals

citations
Cited by 3 publications
(8 citation statements)
references
References 15 publications
1
7
0
Order By: Relevance
“…Based on the precision and recall, we then computed the F-measure as detailed below: 100 1 3 96 0 2 Leaflet 22 0 0 22 0 3 Reddit 74 1 1 72 0 4 CocoaPods 100 0 3 97 0 5 Puma 81 0 6 75 0 6 AutoMapper 68 0 1 67 0 7 MonoDevelop 100 29 8 63 0 8 CodeHub 42 0 8 34 0 9 Manos 100 0 2 98 0 10 puppet 100 0 18 82 0 The '#' symbol gained an F-measure of 0.726, the 'Fix' keyword 0.582 and the 'Bug' keyword only 0.019. Since the F-measure is often used, in the context of information retrieval, to assess the performance of searches, this further test confirms the earlier findings reported by [17], [16]. Analysing the unstructured data of the development logs of the Brackets project as a pilot study, we conclude that the most precise proxy of bug IDs is the '#' identifier, when considering the free-text descriptions of changes written by developers as an addendum to their commits to the VC systems.…”
Section: Evaluating the Precision Of Each Szz Componentsupporting
confidence: 88%
See 2 more Smart Citations
“…Based on the precision and recall, we then computed the F-measure as detailed below: 100 1 3 96 0 2 Leaflet 22 0 0 22 0 3 Reddit 74 1 1 72 0 4 CocoaPods 100 0 3 97 0 5 Puma 81 0 6 75 0 6 AutoMapper 68 0 1 67 0 7 MonoDevelop 100 29 8 63 0 8 CodeHub 42 0 8 34 0 9 Manos 100 0 2 98 0 10 puppet 100 0 18 82 0 The '#' symbol gained an F-measure of 0.726, the 'Fix' keyword 0.582 and the 'Bug' keyword only 0.019. Since the F-measure is often used, in the context of information retrieval, to assess the performance of searches, this further test confirms the earlier findings reported by [17], [16]. Analysing the unstructured data of the development logs of the Brackets project as a pilot study, we conclude that the most precise proxy of bug IDs is the '#' identifier, when considering the free-text descriptions of changes written by developers as an addendum to their commits to the VC systems.…”
Section: Evaluating the Precision Of Each Szz Componentsupporting
confidence: 88%
“…After manual inspection, we found that 40 of the sampled OSS projects were empty. Hence giving an overall number of 'alive' projects of 344 in which we analysed their development logs and bug data automatically using our approach and the proposed tool-chain developed for this research [16]. The data sets (i.e., 344 OSS projects repositories) can be found on Figshare.…”
Section: Project Samplingmentioning
confidence: 99%
See 1 more Smart Citation
“…[37], [38], [44]. Researchers have discussed the lack of integration between revision control systems and issue tracking systems, which affects the recovery of trace links, as well as the prediction of software faults [16], [51], [52]. Researchers addressed that issues form a complex network by themselves, but there were no approaches to predict link types [40], [43].…”
Section: ) Informationmentioning
confidence: 99%
“…A total of 18 studies linked issue reports to commits. [16], [25], [27], [29], [34]- [38], [42], [44], [48], [51], [54]- [56], [58], [62]. Among them, eleven studies focused on recovering the missing trace links between issue reports and commits [16], [27], [29], [36]- [38], [44], [51], [56], [58], [62].…”
Section: ) Issue Reports and Commitsmentioning
confidence: 99%