The Key Technology’s Research Of Trustworthiness Evaluation In Open Source Software Environment

With the rapid development of information technology, Software has become a powerful driving force to support the progress of modernization and civilization. In nowadays, more and more people is paying attention to find the solution to high-quality, low-risk, predictable and controllable software, known as trusted software. Researchers has launched a series of studies on "software trustworthiness". Among those studies, Trustworthiness Evaluation is an important part and the key to application.There are still lots of problems on trustworthiness evaluation. First of all, the theoretical concepts is not uniformed. Scholars tried different methods in different fields. With more models built, more semantic and structural differences were found. The results can’t even be shared. Second, most of the theory is lack of support from real data. Due to the traditional software development method, the developer and application range is so limited that little data could be collected. Lots of the models haven’t link theory with practice.In view of the problems above, starting from the open-source-software, we promoted a general trustworthiness evaluation meta-model and customized it into the OSS environment. We use real data gather from OSS to carry out a series of experiments. We has provided a common solution and reference content to trustworthiness evaluation.Firstly, we use meta-model techniques to promote a domain-independent trustworthiness evaluation meta-model. We has defined the modeling elements within the meta-model. This trustworthiness evaluation meta-model can provided a good solution to the problems on ill-defined content. We abstract seven application steps of the evaluation progress.Secondly, we customized trustworthiness evaluation meta-model in OSS environment. We constructed a detailed design of trustworthy attribute framework (OSS-TAF) and trustworthy evidence framework (OSS-TEF). Base on the two frameworks, we use evidence-attribute mapping method to build a trustworthy software evaluator system (OSS-TSES). We also design the individual indicators algorithms and integrated evaluation algorithm.Thirdly, we use a large number of OSS data obtain from SF.net to apply it into the study of score sheet of OSS-TSES. We carried out a detailed analysis and study of the corresponding data. We has provided provide practical support and demonstration for trustworthiness evaluation.Finally, we write code to achieve an evaluation system. We used a real software as an illustration, evaluated it and draw the results. Through these studies, we has proved the feasibility, accuracy and convenience of the OSS-TSES and the related research.The experiments show that this study will beneficial the trustworthiness evaluation.