Geunseok Yang

Towards Semi-automatic Bug Triage and Severity Prediction Based on Topic Model and Multi-feature of Bug Reports

Bug fixing is an essential activity in the software maintenance, because most of the software systems have unavoidable defects. When new bugs are submitted, triagers have to find and assign appropriate developers to fix the bugs. However, if the bugs are at first assigned to inappropriate developers, they may later have to be reassigned to other developers. That increases the time and cost for fixing bugs. Therefore, finding appropriate developers becomes a key to bug resolution. When triagers assign a new bug report, it is necessary to decide how quickly the bug report should be addressed. Thus, the bug severity is an important factor in bug fixing. In this paper, we propose a novel method for the bug triage and bug severity prediction. First, we extract topic(s) from historical bug reports in the bug repository and find bug reports related to each topic. When a new bug report arrives, we decide the topic(s) to which the report belongs. Then we utilize multi-feature to identify corresponding reports that have the same multi-feature (e.g., Component, product, priority and severity) with the new bug report. Thus, given a new bug report, we are able to recommend the most appropriate developer to fix each bug and predict its severity. To evaluate our approach, we not only measured the effectiveness of our study by using about 30,000 golden bug reports extracted from three open source projects (Eclipse, Mozilla, and Net beans), but also compared some related studies. The results show that our approach is likely to effectively recommend the appropriate developer to fix the given bug and predict its severity.

Predicting severity of bug report by mining bug repository with concept profile

Recently, for large scale software projects, developers rely on bug reports for corrective software maintenance. The severity of a reported bug is an important feature to decide how fast it needs to be fixed. Therefore, to arrange a new submitted bug to an appropriate fixer, it is necessary to recognize the severity of each bug report. Unfortunately, reporters need to decide the severity of bugs manually. Even if there are guidelines on how to verify the severity of a bug, it is still a time-consuming work. Utilizing the concept profiles by mining bug repositories is a good way to resolve this problem. In this paper, we propose a concept profile-based prediction technique to assign the severity of a given bug. In detail, we analyze historical bug reports in the bug repositories and build the concept profiles from them. We evaluate the performance of our method on the bug reports from the bug repositories of popular open-source projects that include Eclipse and Mozilla Firefox, the result shows that the proposed technique can effectively predict the severity of a given bug.

Guiding Bug Triage through Developer Analysis in Bug Reports

An important part of software maintenance is bug report analysis during bug-fixing, especially for large-scale software projects. Since bugs reported to the bug repository need to be fixed, triagers are responsible to identify appropriate developers to execute the fix. Previous research focused on optimizing this process, such as by duplicate detection and use of developer recommendations for reducing the workload of triagers. However, there were scant studies that analyzed developer roles (e.g. reporter and assignee) in the bug-fixing process. Therefore, in this paper, we perform an in-depth empirical study of the different roles that developers perform in bug resolution. By extracting the factors that affect bug resolution from the analysis results, we propose a novel bug triage algorithm to recommend the appropriate developers to fix a given bug. We implement the proposed recommendations on the Eclipse and Mozilla Firefox projects, with the results showing that the new bug triage algorithm can effectively recommend which experts should fix given bugs.

Applying Genetic Programming with Similar Bug Fix Information to Automatic Fault Repair

Owing to the high complexity of recent software products, developers cannot avoid major/minor mistakes, and software bugs are generated during the software development process. When developers manually modify a program source code using bug descriptions to fix bugs, their daily workloads and costs increase. Therefore, we need a way to reduce their workloads and costs. In this paper, we propose a novel automatic fault repair method by using similar bug fix information based on genetic programming (GP). First, we searched for similar buggy source codes related to the new given buggy code, and then we searched for a fixed the buggy code related to the most similar source code. Next, we transformed the fixed code into abstract syntax trees for applying GP and generated the candidate program patches. In this step, we verified the candidate patches by using a fitness function based on given test cases to determine whether the patch was valid or not. Finally, we produced program patches to fix the new given buggy code.

Toward Providing Automatic Program Repair by Utilizing Topic-Based Code Block Similarity

In this paper, we propose the model for automated repair in software fault. Automated patch generation is the most important technique in these days. Genetic Programming (GP) technique is used for automatic program repair, but most of the techniques use just a source code including fault to make initial population. We propose two methods to select similar bug fixing history; using topic modeling and finding similar bugs by using code block similarity.