Software bugs are inevitable and bug fixing is an essential and costly phase during software development. Such defects are often reported in bug reports which are stored in an issue tracking system, or bug repository. Such reports need to be assigned to the most appropriate developers who will eventually fix the issue/bug reported. This process is often called Bug Triaging.

Manual bug triaging is a difficult, expensive, and lengthy process, since it needs the bug triager to manually read, analyze, and assign bug fixers for each newly reported bug. Triagers can become overwhelmed by the number of reports added to the repository. Time and efforts spent into triaging typically diverts valuable resources away from the improvement of the product to the managing of the development process.

To assist triagers and improve the bug triaging efficiency and reduce its cost, this thesis proposes Bugzie, a novel approach for automatic bug triaging based on fuzzy set and cachebased modeling of the bug-fixing capability of developers. Our evaluation results on seven large-scale subject systems show that Bugzie achieves significantly higher levels of efficiency and correctness than existing state-of-the-art approaches. In these subject projects, Bugzie's accuracy for top-1 and top-5 recommendations is higher than those of the second best approach from 4-15% and 6-31%, respectively as Bugzie's top-1 and top-5 recommendation accuracy is generally in the range of 31-51% and 70-83%, respectively. Importantly, existing approaches take from hours to days (even almost a month) to finish training as well as predicting, while in Bugzie, training time is from tens of minutes to an hour.