GETTING ‘WHOLE PICTURE’ BEHAVIOR IN A USE CASEMODEL

摘要

Although widely used, traditional use case modeling does not provide explicit means which couldrnbe easily used for capturing and testing behavior compliance of the entities involved in a particularrnuse case model. Specifically, a use case model (a set of use cases) related to a system under designrnprovides neither an explicit abstraction to capture the “whole picture” of the behavior of thernsystem, nor to cover the interactions of subsystems and internal actors with the parent system. Withrnthe aim to allow for reasoning on the behavior, the paper introduces a simple formal modelrnGeneric UC View which identifies important abstractions and the relations upon them whichrntarget the goal. Among them, the concept of use case expression is the base for the desiredrnreasoning on whether the behavior of an entity (such as an agent, a subsystem or a softwarerncomponent) complies with the composed behavior of its sub-entities, and the behavior on therncommunication links of two neighboring entities is compliant. As a proof of the concept, anrninstance of use case expressions, pro-cases is introduced. Based on behavior protocols, pro-casesrncan be checked for compliance via an already existing verifier. As pro-cases’ syntax is simple,rnresembling regular-expressions, there are simple guidelines for transforming a use case written inrnclassical textual form (based on a template) into a pro-case.