An empirical process to derive OSS defect estimation models

Quality management in Open Source Software (OSS) has become a heated topic since the open source development model emerged. Much work has been done on exploring the distinct quality attributes in OSS, but very few studies covered quality estimation. In this paper, a general procedure is proposed to derive software quality estimation models for OSS projects and various candidate techniques are suggested for individual steps. The purpose is to build a model that estimates the number of defects in a project. Several statistical techniques and a machine learning approach are used to examine the significance of quality predictors. Moreover, a neuro-fuzzy approach is adopted to improve accuracy of the estimation model. This procedure is followed and validated based on data from OSS projects.