When should I use Decision Tables?

A Decision Table is an alternative knowledge representation to Decision Trees which can be used to represent decision lookup when this is the easier and more natural way of expressing the logic.

The "Truth Table" option can be used to generate all the combinations of list Question/Attribute to make it easier for the subject matter expert to then simply assign the values of the outcome column.

Decision Tables can also be used as a vehicle for facilitating the elicitation of the rules-by-example from subject matter experts. Here, each row represents an example or scenario of the decision making logic through mapping the values of the Questions/Attributes columns to the Outcome column value.

Decision Tables also offer a "Decision Tree Induction" option for (List/Boolean outcomes only). This can be used to which automatically generate a Decision Tree representing the equivalent logic of the current set of "examples" (rows) in the Decision Table. This induction process helps to identify any gaps or clashes in the Decision Tables logic which the expert can then rectify with further or better examples. This form of Knowledge Engineering helps to distil the rules from examples provided by domain experts.