A Fuzzy-Based Method for Evaluating the Trustworthiness of Software Processes

摘要

Software trustworthiness is built in its development and maintenance processes. Effective software measurement system can help to assess and predict the trustworthiness of end products in early software lifecycles in order to plan for proper corrective actions to meet trustworthiness objectives. However, it is a challenging task to aggregate individual metric data and derive a single, overall trustworthiness indicator given the hierarchies in software measurement system as well as heterogeneity within a mix of metric data. In addition, very few metric data can be collected precisely and without subjective judgment inherent in data reporters or collecting tools. In this paper, based on the theory of fuzzy set and the AHP method, we propose an evaluation method supporting fuzzy data and multiple types of values, for evaluating process trustworthiness in the form of user-customized levels. A case study is also presented to illustrate the application of our method.