Working with Rule Sets
Within a rule project, you can group several logically related decision entities (such as decision tables, decision trees or event rules) in a rule set. This is useful in cases where two or more decision entities must interact with each other, and the results from one decision entity will be used as input for another decision entity.
Example
If you want to create logically-related decision entities that deal only with order processing, you can create an order processing rule set and use it as a container for all decision entities that apply to order processing.
About Rule Inference
Rule execution is based on making inferences. This means that you can draw a conclusion from a given information with the help of a rule.
In rule execution, it will often be necessary to make inferences over several steps. This means that you use the conclusion drawn from one rule (the result) as input information (the condition) for a second rule. This is called forward chaining. The following two rules illustrate this:
Example
Rule 1: | IF a customer's annual order value is equal to or is larger than $ 5,000, THEN this customer is a VIP customer. |
Rule 2: | IF a customer is a VIP customer, THEN he/she will receive a bonus at the end of a year. |
If you know that a customer's annual order value equals $ 6,000, then you can infer from Rule 1 and Rule 2 that the customer is a VIP customer and will receive a bonus at the end of the year.
This kind of multi-step inferencing can be achieved if you group logically connected decision entities in a rule set and execute it.
Master Rule Set
Besides that, there is a master rule set that contains all of the decision entities that you created in a specific rule project with the exception of decision trees. You cannot add decision entities to or remove decision entities from the master rule set.