Stable operation of computer equipment is important for its effective use. A significant role in ensuring stability depends on the reliability of the software. The article analyzes the state of the problem to improve the reliability of the software based on the models of reliability assessment. The methods of increasing the software reliability using time and software redundancy are analyzed. It is concluded that there is no universal solution to the problem of software reliability in the presence of many private solutions.

Keywords: software reliability, software reliability evaluation models, classification of software reliability models, methods of software reliability improvement

Currently, the software market offers a multitude of diverse DLP-systems designed to protect confidential information from leaks. Since such systems have a relatively high cost, it is advisable to select the optimal option before acquisition and implementation, as incorrect selection will lead to unnecessary costs and ineffective control of information. The paper describes the base of the rules of the fuzzy inference system, designed to evaluate the DLP-system. Practical use of the proposed fuzzy system allows to perform a reasoned choice of such a DLP-system that most fully meets the requirements of a particular organization and provides the required level of protection of information from leaks. At the same time, the evaluation is performed according to several criteria, the values ​​of which, based on expert estimates, are input variables of the fuzzy inference system. At the output of the system of fuzzy inference, a comprehensive evaluation of the DLP-system is formed, which reflects the degree of its suitability for implementation at a particular enterprise.

Keywords: Information security, data protection, protection of confidential information, DLP-system, artificial intelligence, fuzzy logic, fuzzy inference system