Rule flow logic verification: A simulation based approach

In complex rule flow logic verification, simulation based testing approach is one feasible way which tries to go through all the feasible paths. As the baseline, the total number of feasible paths is usually desired to know although it is usually a huge number. The calculation algorithm of feasible paths in a rule flow is proposed in this paper. Using air cargo loading planning rule as an example, we also show how the coverage rate of stochastic algorithm drops sharply as the simulation going on. From rule topologic structure, we propose branch risk measure to denote the impact of the logic error in the branch. This paper also explores to locate the source of logic error through branch tagging and failure test case clustering. The technologies are illustrated in the load planning rule scenario.