Software Trustworthy Evaluation Method Based BP Network

With the widespread use of computer and rapid development of network technology, the software has penetrated into the cultural, economic, political, military and social life at all levels. Software as information infrastructure of the soul plays a very crucial role in all aspects of society. However, as the sale of software is getting larger and the complexity is higher, software is increasingly fragile, and people’s work and life has brought a bad influence, even caused huge losses to life and property. Therefore, as software industry is a national strategic industry, software trustworthy problems naturally become a common concern of the international community.And how to determine the trustworthiness of a software and how to measure the trustworthiness of the software is a main research content of the software trustworthiness evaluation This article studies the software trustworthiness evaluation methods, from the user’s point of view, has the main theoretical and practical significance. This paper mainly conducts the following studies:Firstly, discusses the trustworthiness of the software, trustworthy attributes definition, focuses on descriptions of a variety of software trustworthy attributes, gives each one trustworthy attribute the clear and accurate definition, and on this basis, gives trustworthy attributes classifications and measurement methodsSecondly, Software trustworthiness mutual influence between the trustworthy attributes increases the difficulty of software trustworthy evaluation, resulting not evaluating accurately by traditional methods. To address this problem, proposes a software trustworthiness evaluation model based on BP neural network. This article can assess the trustworthiness of the model by non-linear processing capacity of BP network software effectively.Finally, the proposed method is applied to the "online banking system", based on previous sample data, Web software trustworthy evaluation model is trained by BP neural network, then the model is tested to verify the the BP network model valid and finally t use the BP network model online to evaluate the online banking system.