摘要：Extended Finite State Machine modeling is a widely used technique to model state-based systems. Although EFSM models are usually mainly used to simplify the design and implementation of the systems, their use can be extended to enhance and speed up system maintenance (e.g. Error localization, performance enhancement, change management, etc.). In this paper we present a classification approach for EFSM transitions based on their criticality during maintenance. The purpose of this classification is to give the system maintenance team a tool for estimating criticality level for each transition in the EFSM model and consequently to allow them to better plan and manage the change process according the identified criticality of the transitions involved in the required change. Our classification approach is based on transitions’ complexity as well as the dependencies between the transitions in the model. An empirical study shows that the classification can be used to enhance and speed up the maintenance process for a required change.
其他关键词：Extended finite state machine, system modeling, state-based modeling, transition classification, criticality analysis.
