Skip to content

paulinaaniola/DefectBacklogPrediction

Repository files navigation

DefectBacklogPrediction

Master's thesis project conducted in Software Enginerring faculty

Data

Defect reports comes from open source projects and were collected using Bugzilla bugtracker.

Component Bug tracking period
Eclipse Platform 10-10-2001 - 01-01-2019
Eclipse Birt 15-03-2005 - 01-01-2019
Eclipse Jdt 03-03-2005 - 01-01-2019
Eclipse Data tools 03-03-2005 - 01-01-2019
Eclipse PDE 20-11-2001 - 01-01-2019
Mozilla Firefox 30-07-1999 - 01-01-2019
Mozilla Core 28-03-1997 - 01-01-2019
Mozilla Thunderbird 02-01-2000 - 01-01-2019
Mozilla Calendar 09-11-2000 - 01-01-2019
Kernel File System 18-11-2002 - 01-01-2019
Kernel Networking 15-11-2005 - 01-01-2019
Kernel IO Storage 14-11-2002 - 01-01-2019
Open Office Writer 30-10-2000 - 01-01-2019
Open Office Calc 23-10-2000 - 01-01-2019
Open Office Draw 30-10-2000 - 01-01-2019
Apache Ant 11-09-2000 - 01-01-2019
Apache Apache2 15-01-2001 - 01-01-2019
Libre Office Writer 15-01-2001 - 01-01-2019
Libre Office Calc 08-10-2010 - 01-01-2019
Libre Office Draw 15-01-2011 - 01-01-2019

Data related to the single component consists of three files:

  • component_name_reports - defect reports from Bugzilla
  • component_name_backlog - defect reports converted into weekly defect backlog
  • component_name_samples - set of week numbers from defect backlog for further predictions

These files are located in Data/project_name/component_name folder.

Defect backlog

Defect reports were converted into weekly defect backlog. Every week in the backlog includes the following metrics:

  • backlog_all - the number of all defects in the backlog

  • inflow_all - the number of defects that were reported in corresponding week

  • outflow_all - the number of defects that were resolved in the corresponding week

  • backlog_normal - the number of defects labeled as normal in the backlog

  • inflow_normal - the number of defects labeled as normal that were reported in corresponding week

  • outflow_normal - the number of defects labeled as normal that were resolved in the corresponding week

  • backlog_major - the number of labeled as major defects in the backlog

  • inflow_major - the number of defects labeled as major that were reported in corresponding week

  • outflow_major - the number of defects labeled as major that were resolved in the corresponding week

Samples

For each defect backlog one week was drawn for each month. Week numbers which were drawn are saved in the file component_name_samples. The numbers refers to the week number in the corresponding defect backlog.

About

Master's thesis project in Software Enginerring faculty

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published