UDC 004.7
FEATURES OF USING BAYESIAN GRAPHICAL MODELS
FOR SOFTWARE TESTING
M. M. Zozulya, Researcher, Voronezh State Technical University, Voronezh, Russia;
orcid.org/0000-0003-3929-7254
The article describes an approach to the problem of software testing. The approach is based on Bayesian
graphical models and presents formal mechanisms for logical structuring of software testing problem,
probabilistic and statistical processing of uncertainties that need to be eliminated, the process of developing
and analyzing tests, as well as the inclusion and application of test results. The aim of the work is to provide
a mechanism for dynamic representations of software testing problem based on the construction of models.
They can be used to develop tests, answer "what if" questions, and provide decision support to managers and
testers. The models reflect the knowledge of software tester for further use. The experience of using this approach
in case studies is briefly discussed.
Key words: bayesian graphical models, expert assessments, knowledge gathering, software reliability,
software testing, statistical methods, test development.