what are the various ways to measure performance of production rules
engines? i.e. what are the various factors or metrics that are used to
measure the performance of a forward-chaining inferencing engine in
terms of facts asserted, rules executed, size of rule-sets with
respect to processor speed and memory - i am intersted in obtaining
diffrent scenarios and possibly some numbers as to what is a good
perfromance estimate. Please note that I am not intersted in
benchmarks like the Manners benchmark for rules engines.