An extraction method to collect data on defects and effort evolution in a constantly modified system

Abstract

This paper describes a data extraction method that was carried out on a set of historical development documentation, related to a commercial software system for mobile platform. This method is part of a major project, which aims to identify evidences of technical debt via the analysis of the defect evolution and effort estimation deviation, verifying if there are relations between these concepts and project decisions during the cycles of development. We intend that a future analysis of such data supports the identification of patterns regarding specific decisions and variations of the defect number/frequency and effort deviation. Thus, such patterns could assist project managers during future project decisions, mainly regarding the maintenance and evolution stages.