Metamodel-based Trustworthiness Evaluation Method And Implementation

With the widespread of software applications in various industries, software trustworthiness has become a top concern and also a research hot spot. Software trustworthiness evaluation is a fundamental issue in the study, which is of prime importance in ensuring software trustworthiness.?Given the diversity and differences of evaluation candidates (models, products, documents),? the evaluation need to be performed in a unified and integrated way.The main focus of evaluation models research is the mathematical modeling of evaluation algorithm, and various models have been estalished for different domains. However, those models lack clear definition and description for their structure and semantics. As a result, different models are restricted to their own application patterns; also, it is impossible to share information among models. To solve the above problem, a proposal for the integrated management of software trustworthiness evaluation is presented in this paper.Firstly, based on analysis of current evaluation models, a trustworthiness evaluation meta-model is given by extracting and analyzing basic concepts involved in the general process of trustworthiness evaluation. The consistency of concepts in the meta-model is ensured by using web ontology language OWL to describe semantic information of it.Secondly, using the given meta-model, the creation methods of domain models and evaluation implementation process are discussed. A practical example in the tax industry is explained in details, which shows the process of creating an evaluation model for business applications, and using it to guideline concrete implementations.Finally, a complete solution for a trustworthiness evaluation management system which supports model customization is designed and constructed using J2EE and utilizing B/S (browser/server) structure.?This management system can be used as an assistant in customizing evaluation models, collecting substantiated information, and accomplishing an effective evaluation task.Experiments show that it will be beneficial for successful trustworthiness evaluation practices.