Final Thesis: Towards a Representative and Diverse Analysis of Issue-Tracker Related Code and Process Metrics
Abstract: Issue tracking systems of open source software projects deliver manifold insights to the regarding projects. Due to the huge diversity in the format and structure of the different issue tracker product’s data, performing a large-scale empirical analysis to derive these insights is not a trivial task. We present a uniform integrated data model that is capable of storing data from various issue tracker products. Based on that, we developed a toolkit for crawling and incrementally updating real world issue tracking data into this model. We identified eight key information needs stakeholders of open source software projects have regarding the data from issue tracking systems. Most of these information needs can be satisfied by a set of visualizations and metrics presented within the thesis. Based on data we collected from 13228 projects, we perform an empirical analysis on the structure of the user base. Also, we demonstrate that release dates have no significant and immediate effect on the activity within issue tracking systems. We discuss multiple correlations between trends in the issue tracking systems’ data.
Keywords: Bug trackers, issue trackers, open source analytics, code metrics, process metrics
PDFs: Master Thesis, Work Description
Reference: Maximilian Capraro. Towards a Representative and Diverse Analysis of Issue-Tracker Related Code and Process Metrics. Master Thesis, Friedrich-Alexander-Universität Erlangen-Nürnberg: 2013.