Fuzzing Process Modelling for Web Applications Robustness Testing via Dynamic Bayesian Networks

Maintaining sustainability of the web applications in the conditions of various security threats plays a crucial role in the development of modern information technologies for industrial enterprises, financial institutions and service organizations. This explains the high relevance of developing new science-based efficient computational methods, algorithms and problem-oriented programs for managing stochastic testing processes, applied to web applications with a complex functional structure, internal and external interactions, that tune to their structure and analyze the stochastic flow of information as a result of cross stages dynamic testing. This allows one to analyze and optimize the management process to achieve the test objectives. The article describes the concept of testing processes modeling, research of the obtained models and developing forecasting algorithms, based on the formalized apparatus of dynamic Bayesian networks. The Bayesian models proposed in the research, built on the basis of statistical training, allow formalizing the logical-probabilistic relationships between the elements of the testing process, providing an opportunity to predict test results by the simulation methods of probabilistic inference.