ӘЛЕУМЕТТІК ЖЕЛІЛЕРДЕГІ ГРАФИКАЛЫҚ РЕСУРСТАРДЫ ӨҢДЕУ МОДЕЛІ

Abstract

The advancement of information technologies and social media has led
to an increased risk of extremist content dissemination, which poses a significant threat to
public safety. This is particularly true for visual content (graphics, photos, videos), which is
widely used to disseminate radical ideologies. Effective control requires the use of modern
technologies, particularly machine learning and neural network methods. This article focuses
on investigating the stages of preprocessing graphical content to improve the recognition
of extremist materials using neural network technologies. Data preparation, including
noise removal, image normalization, and feature extraction, plays a key role in enhancing
the accuracy of algorithm performance. Image processing methods such as filtering and segmentation
are discussed, along with their impact on neural network training outcomes. The
proposed approach improves the detection of extremist content on social networks, contributing
to more effective responses to security threats. Experiments have shown that the use
of advanced image processing methods combined with neural network models significantly
increases the accuracy and reliability of recognizing extremist materials.
Keywords: neural network analysis, graphical content, political