This data set consists of 

1 Data set for learning noun-group patterns (from core products) 

2 Data sets for the experiment assemble dependent bug reports (from Firefix product)