On the Effectiveness of Labeled Latent Dirichlet Allocation in Automatic Bug-Report Categorization

Bug-reports are valuable sources of information. However, study of the bug-reports' content written in natural language demands tedious human efforts for manual interpretation. This difficulty limits the scale of empirical studies, which rely on interpretation and categorization of bug-reports. In this work, we investigate the effectiveness of Labeled Latent Dirichlet Allocation (LLDA) in automatic classification of bug-reports into a predefined set of categories.