| Tweet | |
| M. Kondo, D. German, O. Mizuno, and E. Choi, "The Impact of Context Metrics on Just-In-Time Defect Prediction," Empirical Software Engineering, 25(1), 890–939, 2020. | |
| ID | 783 |
| 分類 | 学術論文誌(査読付) |
| タグ | context defect impact just-in-time metrics prediction major-mizuno |
| 表題 (title) |
The Impact of Context Metrics on Just-In-Time Defect Prediction |
| 表題 (英文) |
|
| 著者名 (author) |
Masanari Kondo,Daniel German,Osamu Mizuno,Eun-Hye Choi |
| 英文著者名 (author) |
Masanari Kondo,Daniel German,Osamu Mizuno,Eun-Hye Choi |
| キー (key) |
Masanari Kondo,Daniel German,Osamu Mizuno,Eun-Hye Choi |
| 定期刊行物名 (journal) |
Empirical Software Engineering |
| 定期刊行物名 (英文) |
|
| 巻数 (volume) |
25 |
| 号数 (number) |
1 |
| ページ範囲 (pages) |
890–939 |
| 刊行月 (month) |
0 |
| 出版年 (year) |
2020 |
| Impact Factor (JCR) |
4.457 (2019) |
| URL |
https://doi.org/10.1007/s10664-019-09736-3 |
| 付加情報 (note) |
|
| 注釈 (annote) |
|
| 内容梗概 (abstract) |
Traditional just-in-time defect prediction approaches have been using changed lines of software to predict defective-changes in software development. However, they disregard information around the changed lines. Our main hypothesis is that such information has an impact on the likelihood that the change is defective. To take advantage of this infor- mation in defect prediction, we consider n-lines (n = 1, 2, . . . ) that precede and follow the changed lines (which we call context lines), and propose metrics that measure them, which we call “Context Metrics.” Specifically, these context metrics are defined as the num- ber of words/keywords in the context lines. In a large-scale empirical study using six open source software projects, we compare the performance of using our context metrics, tradi- tional code churn metrics (e.g., the number of modified subsystems), our extended context metrics which measure not only context lines but also changed lines, and combination met- rics that use two extended context metrics at a prediction model for defect prediction. The results show that context metrics that consider the context lines of added-lines achieve the best median value in all cases in terms of a statistical test. Moreover, using few number of context lines is suitable for context metric that considers words, and using more number of context lines is suitable for context metric that considers keywords. Finally, the
combination metrics of two extended context metrics significantly outperform all studied metrics in all studied projects w. r. t. the area under the receiver operation characteristic curve (AUC) and Matthews correlation coefficient (MCC). |
| 論文電子ファイル | 利用できません． |
| BiBTeXエントリ |
@article{id783,
title = {The Impact of Context Metrics on Just-In-Time Defect Prediction},
author = {Masanari Kondo and Daniel German and Osamu Mizuno and Eun-Hye Choi},
journal = {Empirical Software Engineering},
volume = {25},
number = {1},
pages = {890–939},
month = {0},
year = {2020},
impactfactor = {4.457 (2019)},
}
|