A new paper entitled “Bottlenecks in software defect prediction implementation in industrial projects” by Jarosław Hryszko (Volvo IT and WUT) and Lech Madeyski has been published in the Foundations of Computing and Decision Sciences journal, vol. 40, pp. 17-33, March 2015, DOI: 10.1515/fcds-2015-0002.
Here is the abstract of the paper:
“Case studies focused on software defect prediction in real, industrial software development projects are extremely rare. We report on dedicated R&D project established in cooperation between Wroclaw University of Technology and one of the leading automotive software development companies to research possibilities of introduction of software defect prediction using an open source, extensible software measurement and defect prediction framework called DePress (Defect Prediction in Software Systems) the authors are involved in. In the first stage of the R&D project, we verified what kind of problems can be encountered. This work summarizes results of that phase.“