Time series analysis for bug number prediction

Monitoring and predicting the increasing or decreasing trend of bug number in a software system is of great importance to both software project managers and software end-users. For software managers, accurate prediction of bug number of a software system will assist them in making timely decisions, such as effort investment and resource allocation. For software end-users, knowing possible bug number of their systems will enable them to take timely actions in coping with loss caused by possible system failures. To accomplish this goal, in this paper, we model the bug number data per month as time series and, use time series analysis algorithms as ARIMA and X12 enhanced ARIMA to predict bug number, in comparison with polynomial regression as the baseline. X12 is the widely used seasonal adjustment algorithm proposed by U.S. Census. The case study based on Debian bug data from March 1996 to August 2009 shows that X12 enhanced ARIMA can achieve the best performance in bug number prediction. Moreover, both ARIMA and X12 enhanced ARIMA outperform the baseline as polynomial regression.