DEVELOPMENT OF AN EXPERT SYSTEM FOR DETECTING CYBER ATTACKS

This project addresses the limitations of current cyber-security techniques in detecting evolving threats such as phishing, malware, and Distributed Denial of Service (DDoS) attacks, which often exploit unpredictable and dynamic patterns. A web-based, rule-based expert system was developed to serve individuals and small to medium enterprises (SMEs), offering an accessible platform for detecting such threats in real-world environments. Built using the Waterfall methodology, the system progressed through defined phases: requirement analysis, system design, and implementation. The backend, powered by the Flask framework, analyzes user inputs such as email content and traffic logs against predefined rules to identify possible threats. MongoDB stores known attack patterns and logs, ensuring efficient performance and scalability for lightweight deployments. The frontend, created with HTML, CSS, and JavaScript, provides a simple and user-friendly interface, offering real-time feedback like “Phishing Detected: Suspicious domain.” The system intentionally excludes detection of complex threats such as brute force attacks due to its rule-based nature and limited scope. Designed for practical use by non-experts, it offers transparency in how threats are detected, empowering users to understand and manage basic cyber risks. Evaluated in controlled simulations, the platform delivers a cost-effective and scalable solution for strengthening cyber resilience in SMEs and personal digital environments.