Open Your Business Rules!                     
Rules-based Operational Decision Services

OpenRules Classic: Decision Model Validation

An important design feature built into OpenRules is the ability to catch as many rules errors as possible at design time rather that at run-time when it is basically too late.  Just like programming code bugs, an error in a single rule can have catastrophic consequences.  OpenRules provides several tools for rules validation.

Rule Validation with Eclipse Plug-In

OpenRules Eclipse Plug-in diagnoses any errors in Excel-files before you deploy or start your OpenRules-based application.  It automatically diagnoses errors in the Excel-files and displays the appropriate error message(s) inside Eclipse views.  OpenRules displays a diagnostic log with error messages that include hyperlinks that may open the proper Excel file with a cursor identifying the cell where the error occurred.

Batch Rule Validator

If you are using a stand-alone version of OpenRules without the Eclipse IDE, you still can validate your Excel-files.  OpenRules provides a special validation module, the batch Rule Validator, that allows you to test all of the xls-files contained in your rule project. 

You may invoke the Rule Validator by executing the standard Ant target "compile" from the provided build.xml file.  It usually executes two targets -- "compile.java" and "compile.xls".   In Windows, you invoke the target "compile" by double-clicking on the file compile.bat.  The appropriate files are included in most standard OpenRules sample projects.

Automatic Validation of  Conflicts and Completeness with Rule Solver

OpenRules can automatically transfer a decision model defined using business-oriented decision tables into a constraint satisfaction problem that can be validated and solved by an off-the-shelf constraint engine. Constraint-based decision engines immediately brought several advantages:
1) Automatic validation of decision models for consistency and completeness. Additionally to finding conflicts inside one decision table, it can find validate decision models to find conflicts among rules included if different decision tables;
2) Instead of forcing a human decision modeler to specify all (!) rules that lead to one and only one decision, it became possible for a subject matter expert to define only those rules which s/he considers important, and let a decision engine within a limited time to find multiple feasible decisions that satisfy the specified rules and let a user choose the best one;
3) If a decision model specifies an optimization objective, it became possible to automatically find the optimal decision that minimizes or maximizes the specified objective.
Read more in the Rule Solver Manual.

Top>