PDL2's execution mechanism is a combination of
rule-based and similarity based reasoning. Indeed, if a
new pair matches an existing rule, that rule's
activation will be highest (i.e., 1). If there is no
matching rule, then various nodes compete and the
winner is the most similar. When nodes are tied,
giving priority to the more specific one enables the
system to recognize exceptions Then, giving priority
to the one with highest priority allows the system to
deal with conflicting defaults. Finally, giving priority
to the one with highest cover count allows the system
to generalize to the most likely (based on available
evidence) output