Sign inSign up
Analysis and prediction of bug duplicates in KDE bug tracking system thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Analysis and prediction of bug duplicates in KDE bug tracking system

Published on Nov 04, 20113411 Views

Gregor Leban

Bug tracking systems (BTS) are systems that allow users of some software to report to developers bugs they encountered while using it. Common problem of BTS are duplicated reports of the same bug. Sin

SiKDD 2011 - Ljubljana

Related categories

Machine LearningText Mining

Chapter list

Analysis and prediction of bug duplicates in KDE bug tracking system00:00
Agenda00:07
Why are we interested in this problem00:23
What is the problem01:04
Why are bug duplicates a serious problem (1)03:25
Why are bug duplicates a serious problem (2)06:36
Importing the data07:29
Predicting if a bug report is a duplicate (1)08:26
Predicting if a bug report is a duplicate (2)10:07
If not classify, what can we do?10:43
Computing similarity between reports11:31
Are duplicated reports really similar?12:43
Results (graphical)13:19
Improving ranking using meta-data (1)14:09
Improving ranking using meta-data (2)16:56
Future work17:51
Distribution of the number of duplicated reports23:45