Improving Feature Location with Diff Set-Based Heuristics

• The expected results are organized by project.
• The natural language query is given in the name of the file.
• The content of each file lists the methods and the fields that contribute to the feature of interest.

• Different combinations of diff set-based heuristics are evaluated for each feature, separately for Dora and Suade (first work sheet).
• Averages are given per project (second work sheet), as well as for all projects together (third work sheet).