There are some definitions applying to rule-based systems which give a
base for the theory [13].
- Redundant rules -- There are two basic possibilities of defining
redundant rules:
- a rule is redundant if there exists another rule with
identical (up to variables renaming) precondition and conclusion,
- a rule is redundant if it succeeds in the same situation as
another rule and both the rules have the same conclusions.
- Subsumption of rules -- We consider here the most general case of
subsumption. A rule subsumes another rule if the following
conditions hold:
- the precondition part of the first rule is weaker (more
general) than the precondition of the subsumed rule,
- the conclusion part of the first rule is stronger (less
general) than the conclusion of the subsumed rule.
- Reduction of rules -- The main idea of reduction of rules is to
minimize the number of rules without influencing the potential
capabilities of the system for inferring new
knowledge.
- Completeness of rule-based systems -- A rule-based system is
complete if there exists at least one rule succeeding for any
possible input situation.
Igor Wojnicki
2005-11-07