APPLICATION OF METRICS TO REQUIREMENT SPECIFICATION BASED ON THE STATE TRANSITION MODEL.

The application of metrics at the early phases of the software life-cycle is a promising area for research. The state transition model (STM) is one of the most popular requirement models in software engineering. Many requirement specification languages are based on this model. We propose metrics for the STM model and show how these metrics can be used. We study ways of clustering states of a requirement specified using the STM model to reduce complexity of the given STM. Metrics may be used to decide whether there is a need for clustering and to choose between alternative clustering strategies. We specify an example using the STM model and derive metrics from the specification. We then present alternative clustering strategies for the STM.