Write a Blog >>
ICSE 2021
Mon 17 May - Sat 5 June 2021

Many researchers assume that, for software analytics, “more data is better”. We write to show that, at least for learning defect predictors, this may not be true.

To demonstrate this, we analyzed hundreds of popular GitHub projects. These projects ran for 84 months and contained 3,728 commits (median values). Across these projects, most of the defects occur very early in their life cycle. Hence, defect predictors learned from the first 150 commits and four months perform just as well as anything else. This means that, at least for the projects studied here, after the first few months, we need not continually update our defect prediction models.

We hope these results inspire other researchers to adopt a “simplicity-first.” approach to their work. Some domains require a complex and data-hungry analysis. But before assuming complexity, it is prudent to check the raw data looking for “short cuts” that can simplify the analysis.

Wed 26 May

Displayed time zone: Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna change

20:40 - 21:40
2.6.3. Defect Prediction: Data Issues and Bug ClassificationTechnical Track / Journal-First Papers at Blended Sessions Room 3 +12h
Chair(s): Federica Sarro University College London
20:40
20m
Full-paper
Early Life Cycle Software Defect Prediction. Why? How?Technical Track
Technical Track
Shrikanth N C North Carolina State University, Suvodeep Majumder North Carolina State University, Tim Menzies North Carolina State University, USA
Pre-print Media Attached
21:00
20m
Paper
On the Time-Based Conclusion Stability of Cross-Project Defect Prediction ModelsJournal-First
Journal-First Papers
Abdul Ali Bangash University of Alberta, Canada, Hareem Sahar University of Alberta, Abram Hindle University of Alberta, Karim Ali University of Alberta
Pre-print Media Attached
21:20
20m
Paper
IoT Bugs and Development ChallengesArtifact ReusableTechnical Track
Technical Track
Amir Makhshari University of British Columbia (UBC), Ali Mesbah University of British Columbia (UBC)
Pre-print Media Attached

Thu 27 May

Displayed time zone: Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna change

08:40 - 09:40
2.6.3. Defect Prediction: Data Issues and Bug ClassificationJournal-First Papers / Technical Track at Blended Sessions Room 3
08:40
20m
Full-paper
Early Life Cycle Software Defect Prediction. Why? How?Technical Track
Technical Track
Shrikanth N C North Carolina State University, Suvodeep Majumder North Carolina State University, Tim Menzies North Carolina State University, USA
Pre-print Media Attached
09:00
20m
Paper
On the Time-Based Conclusion Stability of Cross-Project Defect Prediction ModelsJournal-First
Journal-First Papers
Abdul Ali Bangash University of Alberta, Canada, Hareem Sahar University of Alberta, Abram Hindle University of Alberta, Karim Ali University of Alberta
Pre-print Media Attached
09:20
20m
Paper
IoT Bugs and Development ChallengesArtifact ReusableTechnical Track
Technical Track
Amir Makhshari University of British Columbia (UBC), Ali Mesbah University of British Columbia (UBC)
Pre-print Media Attached