Software components prioritization using OCL formal specification for effective testing

Abstract

In soft real time system development, testing effort minimization is a challenging task. Earlier research has shown that often a small percentage of components are responsible for most of the faults reported at the later stages of software development. Due to the time and other resource constraints, fault-prone components are ignored during testing activity which leads to compromises on software quality. Thus there is a need to identify fault-prone components of the system based on the data collected at the early stages of software development. The major focus of the proposed methodology is to identify and prioritize fault-prone components of the system using its OCL formal specifications. This approach enables testers to distribute more effort on fault-prone components than non fault-prone components of the system. The proposed methodology is illustrated based on three case study applications.