An Improved Comprehensive Evaluation Model of Software Dependability based on Rough Set Theory

Dependability of software, a major concern in many computer applications, can be improved through several means. But systematic approaches for its evaluation do not exist, which is the prerequisite for dependability control and improvement. Software dependability evaluation is an urgent problem to be solved. There is some subjectivity about weighting coefficient when applying fuzzy comprehensive evaluation model to evaluate software dependability. A new comprehensive evaluation model with combinational weight based on rough set theory was proposed in this paper. In this new model, subjective weight was obtained from expert’s judgment by using the traditional fuzzy comprehensive evaluation method, and the objective weight was determined from statistical data by using the rough set theory. Then, a relatively reasonable weight was obtained by combining the subjective weight and the objective weight, which effectively avoids subjectivity of weighing coefficient in fuzzy comprehensive evaluation model. In the example, two different dependability evaluation results were derived from the traditional evaluation model and the improved model respectively. The comparative result proves the feasibility and validity of this improved comprehensive evaluation model.