Global Environment Analysis

Abstract

In the current digital age, web security is under constant
threat from malicious users and evolving cyberattacks.
As web applications and cloud platforms expand, the
attack surface grows significantly, increasing
vulnerability to threats such as SQL injection, cross-site
scripting (XSS), denial-of-service (DoS), and more. It is
crucial to detect and mitigate these attacks efficiently and
in real time. Traditional rule-based intrusion detection
systems (IDS) are insufficient in detecting advanced
persistent threats, as they often rely on static signatures
and patterns. This has led to the growing adoption of
machine learning (ML) and deep learning (DL)
techniques, which offer the potential to learn from past
data and identify complex attack patterns. This research
project presents a comparative study on various ML and
DL models—including SVM, Decision Trees, Random
Forests, CNNs, and RNNs—for the detection of web
attacks. By using benchmark datasets and various
evaluation metrics, the study aims to identify the
strengths and limitations of each method. The outcome of
this research will aid in understanding which models are
more efficient under different conditions and constraints.
It will also offer insights into developing hybrid or
ensemble approaches for real-time, scalable, and accurate
web attack detection systems.