skip to main content
Corticon Studio: Rule Modeling Guide : Logical analysis and optimization : Validating and testing Rulesheets in Corticon Studio : The completeness checker : Limitations of the Completeness Checker
 

Try Corticon Now
Limitations of the Completeness Checker
The Completeness Checker is powerful in its ability to discover missing combinations of Conditions from your Rulesheet. However, it is not smart enough to determine if these combinations make business sense or not. The example in the following figure shows two rules used in a health care scenario to screen for high-risk pregnancies:
Figure 224. Example Prior to Completeness Check
Now, we will click on the Completeness Checker:
Figure 225. Example after Completeness Check
Notice that columns 3-4 have been automatically added to the Rulesheet.  But also notice that column 3 contains an unusual Condition: gender <> female. Because the other two Conditions in column 3 have dash values, we know it contains component or sub-rules. By double-clicking on column 3's header, its sub-rules are revealed:
Figure 226. Non-Female Sub-Rules Revealed
Because our Rulesheet is intended to identify high-risk pregnancies, it would not seem necessary to evaluate non-female (i.e., male) patients at all. And if male patients are evaluated, then we can say with some certainty that the scenarios described by sub-rules 3.1 and 3.3 – those scenarios containing pregnant males – are truly unnecessary.  While these combinations may be members of the Cross Product, they are clearly not combinations that can occur in real life. If other rules in an application prevent combinations like this from occurring, then sub-rules 3.1 and 3.3 may also be unnecessary here. On the other hand, if no other rules catch this faulty combination earlier, then we may want to use this opportunity to raise an error message or take some other action that prompts a re-examination of the input data.