Advanced Security Scanner for Malware Detection and USB Monitoring with Gmail Integration

The aim of the project is to develop a comprehensive malware detection system that enforces a combination of detection mechanisms to provide resilient and scalable defence against a wide-range of cyber threats. The system uses the Virus Total API to detect known malware, accessed through a large collection of signatures provided by various anti-virus engines. In addition to the signature-based detection, the system relies on YARA rule-based identification, which recognises malware by its particular patterns and file characteristics, making its degree of security even higher. With machine learning (ML) in place, the system can identify hitherto unknown malware by analyzing data patterns, like file metadata and system behaviors. In addition, with the inclusion of deep learning methodologies, namely Convolutional Neural Networks (CNNs) we automatically learn complex data features and patterns, contributing towards enhanced detection of state-of-the-art and evasive malware. The graphical malware scanner can also perform on-the-fly scanning of files as they are accessed, and if discrepancies are found – regardless of how the files are accessed such as through a file metilnk or ifa USB/DVD content is inserted to a compromised computer that has autorun enabled – it will block malware located in.exe files as a result of this on-access scan. The service also scans email attachments with the Gmail API, preventing malware from being delivered through email.