Study On Present Situation And Characteristic Analysis Of The Mozilla Bug Repository

The software repository mining has always been an important part of the data mining. It plays a major role in software development, testing and maintenance. The objective of data warehouse mining is to support the decision-making process in the project management. Data warehouse is subject-oriented integrated data collection at different periods. Developers can find the interesting patterns and other hidden knowledge by data mining technics so that they can improve software development efficiency. In order to reduce maintenance costs, project developers should track the defects on software running to get the system current state. Also developers have to fix the bugs as soon as possible so that the software can work well and improve system stability.Bug repository management is one of the most important issues in large-scale open-source software maintenance. This paper collects and analyzes the history information of bug reports in Bugzilla, and shows some fundamental facts of the Mozilla bug repository. We explain the life-cycle of a bug report in detail to help users understand the bug report management better. Secondly, we discuss bug report attributes including status, severity, resolution as well as the process of submission and searching. It can give users a correct knowledge of a bug report. By analyzing the distribution and resolved time of reports, allowing people get a better cognition of the composition of Mozilla bug report database and the management cost.Empirical experiments find that the number of new bug reports increases right after new releases of Mozilla main products. Bug reports with high level severity can be resolved in a relatively short period. There is a negative correlation between the number of bug reports per assignee and the fix rate of that assignee. And after2005, the fix rate increases and the resolved time decreases.This paper provides a general situation of the Mozilla bug repository. It can help users and developers gain a correct understanding of the large-scale open-source software and operate projects better.