Sciweavers

CSMR
2008
IEEE

Trend Analysis and Issue Prediction in Large-Scale Open Source Systems

14 years 6 months ago
Trend Analysis and Issue Prediction in Large-Scale Open Source Systems
Effort to evolve and maintain a software system is likely to vary depending on the amount and frequency of change requests. This paper proposes to model change requests as time series and to rely on time series mathematical framework to analyze and model them. In particular, this paper focuses on the number of new change requests per KLOC and per unit of time. Time series can have a two-fold application: they can be used to forecast future values and to identify trends. Increasing trends can indicate an increase in customer requests for new features or a decrease in the software system quality. A decreasing trend can indicate application stability and maturity, but also a reduced popularity and adoption. The paper reports case studies over about five years for three large open source applications: Eclipse, Mozilla and JBoss. The case studies show the capability of time series to model change request density and provide empirical evidence of an increasing trend in newly opened change ...
Bénédicte Kenmei, Giuliano Antoniol,
Added 29 May 2010
Updated 29 May 2010
Type Conference
Year 2008
Where CSMR
Authors Bénédicte Kenmei, Giuliano Antoniol, Massimiliano Di Penta
Comments (0)