Information and communication and chemical technologies

Vol. 2 No. 27 (2025) / 2025-06-28 / Number of views: 313

ALGORITHM FOR MULTI-FACTORED FORECASTING OF NETWORK VULNERABILITIES: FROM CVE DATA ANALYSIS TO THE DEVELOPMENT OF A FORECASTING MODEL

Authors

L.N. Gumilyov ENU
https://orcid.org/0000-0002-9212-251X
K.Kulazhanov KazUTB
https://orcid.org/0009-0002-5754-4001
K.Kulazhanov KazUTB
https://orcid.org/0000-0002-1195-6919
Technical University of Berlin
https://orcid.org/0009-0002-9480-0596

Keywords

multi-factored forecasting algorithm, network vulnerabilities, forecasting methods, CVSS, CVE, machine learning

Link to DOI:

https://doi.org/10.58805/kazutb.v.2.27-1006

How to quote

Shegetaeva, A., et al. “ALGORITHM FOR MULTI-FACTORED FORECASTING OF NETWORK VULNERABILITIES: FROM CVE DATA ANALYSIS TO THE DEVELOPMENT OF A FORECASTING MODEL”. Vestnik KazUTB, vol. 2, no. 27, June 2025, doi:10.58805/kazutb.v.2.27-1006.

Abstract

This paper presents an algorithm for multi-factored forecasting of network vulnerabilities based on data analysis from the CVE (Common Vulnerabilities and Exposures) database. The goal of the research is to develop and implement a model capable of not only detecting existing vulnerabilities but also predicting their future occurrence using formal methods and machine learning. The authors conduct an overview and comparative analysis of modern forecasting methods, including statistical approaches, machine learning algorithms (specifically Random Forest), clustering, and artificial intelligence methods. As part of the study, a forecasting program was developed, which uses data on vulnerability types, severity levels (CVSS), software, and the year of detection. The model, based on the Random Forest algorithm, demonstrated high forecasting accuracy (94.14%). The results were visualized in the form of graphs and charts, reflecting the dynamics and distribution of vulnerabilities by year and software categories. The results of the work confirm the effectiveness of the proposed approach and demonstrate its potential application in cybersecurity systems for early threat detection. In the future, it is planned to integrate more complex models, account for new factors, and create a system for automatic threat notifications.