Researchers analyzing low-quality software rely on historical data, particularly on when defects were introduced into codebase. The SZZ algorithm and its variants track modified lines in bug-fixing commits to identify bug-introducing changes. However, SZZ struggles with accuracy, especially in cases of unrelated modifications (tangled commits) or missing references to external files (ghost commits) in the version control system. Our research explores whether incorporating bug discussions can improve SZZ by identifying relevant files. Using a dataset of 12,472 Mozilla bug reports (RoTEB dataset), we analyzed a sample of 369 reports and found that files are referenced in discussions for various reasons, such as system dumps, bug description, or solution draft, and they are valuable. Augmenting SZZ with this information improved precision in identifying bug-introducing commits but had little impact on recall. Our findings highlight the potential of bug discussions in enhancing SZZ, paving the way for further refinements. Data: https://zenodo.org/records/11484723
Mon 23 JunDisplayed time zone: Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna change
16:00 - 18:00 | MSR 1Industry Papers / Ideas, Visions and Reflections / Research Papers / Journal First at Aurora B Chair(s): Andrew Begel Carnegie Mellon University | ||
16:00 20mTalk | On Refining the SZZ Algorithm with Bug Discussion Data Journal First Pooja Rani University of Zurich, Fernando Petrulio University of Zurich, Alberto Bacchelli University of Zurich | ||
16:20 20mTalk | SemBIC: Semantic-aware Identification of Bug-inducing Commits Research Papers Xiao Chen The Hong Kong University of Science and Technology, Hengcheng Zhu The Hong Kong University of Science and Technology, Jialun Cao Hong Kong University of Science and Technology, Ming Wen Huazhong University of Science and Technology, Shing-Chi Cheung Hong Kong University of Science and Technology DOI | ||
16:40 20mTalk | Evaluating SZZ Implementations: An Empirical Study on the Linux Kernel Journal First Yunbo Lyu Singapore Management University, Hong Jin Kang University of Sydney, Ratnadira Widyasari Singapore Management University, Singapore, Julia Lawall Inria, David Lo Singapore Management University | ||
17:00 10mTalk | HyperSeq: A Hyper-Adaptive Representation for Predictive Sequencing of States Ideas, Visions and Reflections | ||
17:10 10mTalk | LLMs for Defect Prediction in Evolving Datasets: Emerging Results and Future Directions Ideas, Visions and Reflections Umamaheswara Sharma B National Institute of Technology, Calicut, Farhan Chonari National Institute of Technology Calicut, Gokul K Anilkumar National Institute of Technology Calicut, Saikiran Konchada National Institute of Technology Calicut | ||
17:20 20mTalk | ROSE LCOM Tools Industry Papers Kenneth Lamar University of Central Florida, Peter Pirkelbauer Lawrence Livermore National Laboratory, Zachary Painter University of Central Florida, Damian Dechev University of Central Florida |
Aurora B is the second room in the Aurora wing.
When facing the main Cosmos Hall, access to the Aurora wing is on the right, close to the side entrance of the hotel.