I hadn’t considered ruleflow asa decision flow - mostly it is used to coordinate rulesets - but probably we need to consider in the same sense as decision graphs and trees…

Cheers

Decision model should provide semantics and notation for modeling arbitrary decision processes. It could be used different models to do this following semantics of for eg., UML Activity or UML State Machine. The idea behind this is to split complex decisions into flow of micro decision activities connected with control flow or bind to state transitions (in case of statefull interactive decision model). Micro decision activities could be realized by rule sets, or any other behavioral representation of the rules oriented decision logic, for eg. decision tables, decision trees). Similar approach for future extensions was mentioned in PRR as well as supported by FICO ruleflow implementation. It seems iLog has different semantics of what ruleflow is and scopes it’s application to represent the logic of the single rule set. Although, rule set as well as other types of decision entities are subclasses of Behaviour (according PRR) and equally could be used to represent an arbitrary decision process the restriction to use a single decision table or decision tree for modeling decision process seems to come from the lack of currently available commercial technology support.

Regards,
Gediminas