I have knowledge stored facts that state conditions to be evaluated and
actions that should be taken in those conditions. I know that the previous
description of the facts sounds exactly like rule definition. However, there
would be additional "metarules" that solve whether there exists sufficient
conditions to even try to match those conditions in a particular group of
facts.
Would it be possible to write that kind of "metarules" that would evaluate
an expression stored in a fact? How much would this impair the matching
speed compared to ordinary rules?
I realize that it would be possible to combine the conditions of the
"metarules" with the conditions of the facts to generate new rule that would
only fire when both conditions are satisfied. In practice, this is very
unconvinient because it clutters up those simple "fact-rules" pretty good.
Especially if there are many partially overlapping metarules, this would
lead to multiple new rules with really many conditions.
In addition, it would be impossible to easily modify those
metarule-conditions in many places because same kind of conditions can exist
in both meta-conditions and fact-conditions. Of course, there could be a
rulecompiler that would generate those new rules from that information and
thus avoid the modification problems.
Any suggestions are appreciated.
---------------------------------------------------------------------
To unsubscribe, send the words 'unsubscribe jess-users [EMAIL PROTECTED]'
in the BODY of a message to [EMAIL PROTECTED], NOT to the
list (use your own address!) List problems? Notify [EMAIL PROTECTED]
---------------------------------------------------------------------
