Leveraging Defects Life-Cycle for Labeling Defective Classes

<p>Data from software repositories are a very useful asset to building dierent kinds of</p> <p>models and recommender systems aimed to support software developers. Specically,</p> <p>the identication of likely defect-prone les (i.e., classes in Object-Oriented systems)</p> <p>helps in prioritizing, testing, and analysis activities. This work focuses on automated</p> <p>methods for labeling a class in a version as defective or not. The most used methods</p> <p>for automated class labeling belong to the SZZ family and fail in various circum-</p> <p>stances. Thus, recent studies suggest the use of aect version (AV) as provided by</p> <p>developers and available in the issue tracker such as JIRA. However, in many cir-</p> <p>cumstances, the AV might not be used because it is unavailable or inconsistent. The</p> <p>aim of this study is twofold: 1) to measure the AV availability and consistency in</p> <p>open-source projects, 2) to propose, evaluate, and compare to SZZ, a new method</p> <p>for labeling defective classes which is based on the idea that defects have a stable</p> <p>life-cycle in terms of proportion of versions needed to discover the defect and to x</p> <p>the defect. Results related to 212 open-source projects from the Apache ecosystem,</p> <p>featuring a total of about 125,000 defects, show that the AV cannot be used in the</p> <p>majority (51%) of defects. Therefore, it is important to investigate automated meth-</p> <p>ods for labeling defective classes. Results related to 76 open-source projects from the</p> <p>Apache ecosystem, featuring a total of about 6,250,000 classes that are are aected</p> <p>by 60,000 defects and spread over 4,000 versions and 760,000 commits, show that the</p> <p>proposed method for labeling defective classes is, in average among projects and de-</p> <p>fects, more accurate, in terms of Precision, Kappa, F1 and MCC than all previously</p> <p>proposed SZZ methods. Moreover, the improvement in accuracy from combining SZZ</p> <p>with defects life-cycle information is statistically signicant but practically irrelevant</p> <p>(</p> <p>overall and in average, more accurate via defects' life-cycle than any SZZ method.</p>

Name thesis:degree_name

MS in Computer Science

Discipline thesis:degree_discipline

Computer Science

Year dc:date.available

2019

Author dc:creator

• Vandehei, Bailey R

Contributors dc:contributor

• Davide Falessi
• Computer Science
• College of Engineering

• affect version
• defect prediction
• dataset
• Software Engineering

Identifier

10.15368/theses.2019.123

OAI identifier oai:identifier

oai:digitalcommons.calpoly.edu:theses-3565