Predicting Software Defects Across Projects


Principal Investigator
Burak Turhan
This postdoctoral targeted project was forced for an early completion by the funding organization due to career progress and appointment as Full Professor
Start date
01 September 2012
End Date
31 December 2013 (31 August 2015)
Funding Organization
Academy of Finland
Funding Reference
Funding Amount
€138.762 (€311.420)
Total Budget
€198.231 (€389.279)


  • Prof. Tim Menzies @ West Virginia University
  • Prof. Ayse Bener @ Ryerson University
  • Dr. Andrian Marcus @ Wayne State University
  • Elektrobit Oy
  • Dr. Thomas Zimmermann @ Microsoft Research
  • Dr. Forrest Shull @ Fraunhofer USA

Public Project Description

This project has shown that the extra effort associated with collecting data from other projects is not feasible in terms of practical performance improvement when there is already an established within project defect predictor using full project history. However, when there is limited project history, e.g. early phases of development, mixed project predictions are justifiable as they perform as good as full within project models. The results also indicate that the lessons learned after combining small parts of different data sources were superior to either generalizations formed over all the data or local lessons formed from particular projects. When researchers attempt to draw lessons from some historical data source, they should ignore any existing local divisions into multiple sources, cluster across all available data, then restrict the learning of lessons to the clusters from other sources that are nearest to the test data.