Skip to main content
notice

Seminar by Dongsun Kim (University of Luxembourg)

June 3, 2015
|


Speaker: Dongsun Kim
                University of Luxembourg

Title: Good Hunting: Locating, Prioritizing, and Fixing Bugs Automatically

Date: Wednesday, June 3rd, 2015

Time: 10:30am

Place: EV 2.184

ABSTRACT

Developers are faced with tons of bugs during programming tasks. Dealing with bugs, i.e., debugging, is often tedious and time-consuming. Typical bug resolution tasks include the identification of buggy locations, the selection of bugs to fix, and the actual repair of them. This talk leads audiences to a journey of bug hunting by introducing three novel techniques for locating, prioritizing, and fixing bugs. First, the localization technique builds two-phase recommendation model. This filters out deficient bug reports in the first phase to avoid noisy input and enhance the accuracy of bug localization and, in the second phase, recommends potential buggy locations. The bug prioritization technique utilizes machine learning techniques to select bugs to fix first. This technique first extracts three different groups of features from bug data: history, complexity, and network features. Then, it builds a prediction model based on the features to identify bugs with high priority. The program repair technique leverages human-written patches to automatically generate more acceptable bug patches. Many fix patterns can be identified from human-written patches, the technique uses the patterns to create patch candidates. These are more acceptable to developers than those from random mutation.

BIO

Dongsun Kim is a Research Associate at the University of Luxembourg. He was formerly a post-doctoral fellow at the Hong Kong University of Science and Technology. His research interest includes automatic patch generation, fault localization, static analysis, and search-based software engineering (SBSE). In particular, automated debugging is his current focus. He is also pursuing a topic on change pattern mining for program repair. His recent work has been recognized by several awards such as a featured article of the IEEE Transactions on Software Engineering (TSE) and ACM SIGSOFT Distinguished Paper of the International Conference on Software Engineering (ICSE).




Back to top

© Concordia University