Specically, when the set cardinalities are large, DCJ introduces a
signicant performance improvement as compared to PSJ. On the other hand,
PSJ wins over DCJ when small sets are used. The work presented in [HM97],
[RPNK00], and in this paper raises the question whether even better algorithms
for set containment joins exist. Currently, we are trying to develop a hybrid
algorithm that combines the strengths of PSJ and DCJ. Developing eÆcient al-
gorithms for other set join operators, for instance the intersection join, is another
challenging and mostly unexplored research direction.