it is a value between 0 and 1 that defines the probability that the inference generated from a rule is true
the certainty factor is added to each result from an IF THEN rule (that is, its attached to the 'then' part.)
When multiple inferences are generated, and the result has come from multiple conditions and results, then the certainty factors from each inference are multiplied.
for example:
an inference could have a 90% chance of being true
the second IF-THEN rule ive used has a 10% chance of being true
IF condition1
THEN inference1 (0.9) (90% chance of result being true)
IF condition 2
THEN inference 2 (0.1)
final probability of being true = 9%
(hence the probability of the final result being true is 9% because the two factors are multiplied, 0.9*0.1 = 0.09.)
im guessing a bit here, however. maybe someone can check..