CYBERATTACK PREVENTION THROUGH THE DEVELOPMENT OF A MULTI-VULNERABILITY SCANNER

Abstract

This study presents the development of a user-friendly web-based scanning tool designed to detect cybersecurity threats from three common sources: executable (.exe) files, URLs, and email messages (.eml). The objective of this study is to provide an accessible and effective cybersecurity solution for users of all technical backgrounds by enabling real-time threat detection, persistent scan history tracking, user-generated feedback, and comprehensive analytical reports with preventive recommendations. The methodology involves applying distinct threat detection techniques personalized to each input type and utilizing the Agile Kanban methodology to implement each step of the software design process, enabling flexible adaptation and visual tracking of development progress. For URLs, detection is performed using a machine learning approach that combines an XGBoost classifier with domain reputation assessments to identify malicious links. Email threat analysis is conducted through multi-layered inspection, including header analysis, attachment scanning, and content classification using a Support Vector Machine (SVM) trained on Word2Vec-generated feature vectors. Executable file analysis combines cryptographic hash verification, signature pattern matching, heuristic rule evaluation, and external threat intelligence via the VirusTotal API to identify potential threats. The results of the experiments demonstrate that the system is able to detect a wide range of threats while also maintaining usability and performance, which contributes to greater cybersecurity awareness and resilience for the typical user.