Scenario path identification for distributed systems: A graph based approach

Abstract

With increased complexity of software systems being developed; analysis of use case scenarios is gaining importance leading to effective test case identification during early part of the life cycle. Existing approaches provide various methods for analysis of UML activity diagrams and scenario path identification based on graph models of activity diagrams. In most cases these methods consider a single activity diagram. However use case scenarios may span multiple activity diagrams, which have become quite common with distributed development of software systems. In this paper we propose Activity Relationship graph model that depicts the interrelationship of activity diagrams modeling a use case. Activity Relationship graph ARG is a hierarchical graph where each node depicts an activity diagram modeled as activity diagram graph (AG). We also define a set of metrics named Use case Scenario Paths (USP) that measures the minimum number of independent paths in ARG. An algorithm is proposed to analyze ARG and derive the number of Use case Scenario Paths. This gives a measure of the number of test paths for a requirement based on analysis models early in the life cycle.