Using Decision Tables › Chaining

Chaining

Decision Table-rules fully participate in chaining along with rules of any other type. The engine can use a Decision Table-rule for both forward chaining and backward chaining. Once the engine has fired or failed a Decision Table-rule, the rule no longer participates in chaining, unless reinstated due to Truth Maintenance side-effects.

When backward chaining, the engine selects a Decision Table-rule if any of its Actions can resolve the primary goal or any current subgoals. When firing a Decision Table-rule, the engine evaluates Conditions and performs Actions as it does for forward chaining, that is, the engine does NOT somehow focus on particular Conditions or Actions which are known to contribute to goal resolution. As a result, the rule may fire without contributing to goal resolution. And, as noted above, once fired or failed, the rule no longer participates in chaining.