Network Guard: Enhancing Cybersecurity and Network Efficiency Through Data Insights

With cyber-attacks on the rise, having a strong and well-planned network security system is essential for keeping private information safe and ensuring smooth communication. This paper presents NetworkGuard, a web-based tool designed to improve cybersecurity by watching network traffic in real time and spotting threats using machine learning. Using the KDD Cup 1999 dataset, the system applies Random Forest method to label network data as either harmful or safe, with accuracy rates of 98 - 99%. The backend, built with Node.js, works together with MongoDB to safely store packet information, while the frontend which is created using HTML/CSS and Bootstrap, offers an easy-to-use dashboard that shows live attack data. Hosted on AWS EC2, the platform can grow to handle more users and keeps working even when there's a lot of traffic, giving the administrator useful, real-time updates. Google reCAPTCHA adds another layer of protection by stopping automated login attempts. Experimental test results show the system can successfully identify different types of attacks, such as Denial-of-Service (DoS) and probing, with less than 500 milliseconds of delay for real-time detection. By combining smart machine learning with a responsive web design, NetworkGuard acts as a hands-on defense system against fast-changing cyber threats and sets a new standard for flexible security tools.