SNIFFING DETECTION BASED

ON MACHINE LEARNING
Vasanth P Grihith S Youkesh G
Balamurugan Department of Computer Department of Computer Department of Computer
Department of Computer Science and Engineering Science and Engineering Science and Engineering
Science and Engineering Sri Sairam Engineering Sri Sairam Engineering Sri Sairam Engineering
Sri Sairam Engineering College College College
College Chennai,India Chennai,India Chennai,India
Chennai,India : email address [email protected] Youkeshganesan1103@g
email address m ail.com

Abstract—The rapid growth of open networks and increased

societal dependence on them have given rise to new cyber threats, II. PROBLEM STATEMENT
particularly the alarming prevalence of sniffing attacks. These
malicious activities exploit vulnerabilities in network systems, In today's increasingly interconnected world, societies
tricking unsuspecting users and compromising sensitive heavily rely on open networks, such as the internet, for various
information. The consequences are staggering, with billions of activities. However, this growing dependence has given rise to
dollars lost each year due to these attacks. Therefore, the need to a multitude of cyber threats and information security issues,
secure websites and detect sniffing attacks has become posing significant challenges for individuals and organizations
paramount. In this project, we propose an advanced sniffing alike. One of the most critical and prevalent attacks is the
detection system that combines machine learning algorithms, exploitation of malicious URLs, particularly through sniffing
behavioural analysis, and anomaly detection techniques. Our techniques, aimed at extracting unsolicited information. These
solution operates at both the network and application levels, attacks often target inexperienced end users, tricking them into
providing comprehensive protection against various types of accessing malicious links that compromise their systems and
sniffing techniques. Through deep packet inspection, behavioural lead to substantial financial losses. The impact of such attacks
analysis, and application-level analysis, our system can identify includes risks like data theft, financial fraud, privacy breaches,
and mitigate sniffing attacks in real-timeKey features of our and unauthorized system access. These incidents collectively
solution include real-time alerts, automated response actions, result in billions of dollars lost each year. Given the alarming
seamless integration with existing security infrastructure, and a consequences and the accelerating pace of these threats, it has
user-friendly interface for easy configuration and management. become imperative to enhance website security measures.
We have collected a diverse dataset of 30,000 URLs for training Therefore, the focus of this project is to develop effective
and testing, and our feature extraction module utilizes over 20 techniques for detecting and mitigating malicious URLs. By
user-defined functions to assess the maliciousness of URLs addressing this problem, we aim to safeguard users from falling
accurately. By developing and implementing our sniffing victim to
detection solution, organizations can proactively safeguard their these deceptive tactics, protecting their sensitive information,
systems, prevent financial losses, and protect sensitive data from and reducing the financial losses incurred due to cybercriminal
falling into the wrong hands. Our solution represents a crucial activities..
step in securing open networks, ensuring the integrity of
information, and mitigating the risks associated with sniffing III METHODOLOGY
attacks

Keywords—detection, sniffing, , Dataset, sniffing attacks ,

intergrity

I. INTRODUCTION

In today's increasingly interconnected world, societies heavily rely on

open networks, such as the internet, for various activities. However,
this growing dependence has given rise to a multitude of cyber threats
and information security issues, posing significant challenges for
individuals and organizations alike. One of the most critical and
prevalent attacks is the exploitation of malicious URLs, particularly
through sniffing techniques, aimed at extracting unsolicited
information. These attacks often target inexperienced end users,
tricking them into accessing malicious links that compromise their
III. MODULE DESCRIPTION
systems and lead to substantial financial losses. The impact of such
attacks includes risks like data theft, financial fraud, privacy breaches, After the text edit has been completed, the paper is ready
and unauthorized system access. These incidents collectively result in for the template. Duplicate the template file by using the Save
billions of dollars lost each year. Given the alarming consequences As command, and use the naming convention prescribed by
and the accelerating pace of these threats, it has become imperative to your conference for the name of your paper. In this newly
enhance website security measures. Therefore, the focus of this created file, highlight all of the contents and import your
project is to develop effective techniques for detecting and mitigating prepared text file. You are now ready to style your paper; use
malicious URLs. By addressing this problem, we aim to safeguard the scroll down window on the left of the MS Word Formatting
toolbar.
users from falling victim to these deceptive tactics, protecting their
sensitive information, and reducing the financial losses incurred due
to cybercriminal activities.
 performance.
A. Module1: Dataset Collection
• Using the collected dataset, the system undergoes a training
• In this module, the primary goal is to collect the necessary phase to learn and identify patterns that differentiate
datasets for training and testing the malicious URL detection malicious URLs from benign ones
system.
• Collecting a diverse and representative dataset is crucial for •The training process employs machine learning algorithms or
training an effective model. other classification techniques
The dataset includes a collection of both malicious and benign
URLs. • The trained model is then tested using various evaluation
These URLs are sourced from various reliable and relevant metrics to assess its accuracy, precision, recall, or F1 score.
sources, such as security databases, online forums, or previous
research studies.
• The dataset used for the prototype development consists of
30,000 URLs, providing a substantial amount of data for training
and evaluation.

'IV RESULT

B. Module 2: Feature Extraction

The feature extraction module focuses on extracting
relevant features from the URLs to determine their
maliciousness.
• Various features can help identify patterns and characteristics
commonly found in malicious URLs.
• This module employs more than 20 user-defined functions
that extract different features from the URLs. These functions
may include analyzing the length of the URL, presence of
suspicious characters, domain reputation, use of redirection, or
the presence of specific keywords.
• The extracted features serve as valuable indicators for the
subsequent analysis and classification of URLs as malicious or
benign.

C. Module 3: Training and Testing REFERENCES

The training and testing module involves the core process of 1. Sultan Asiri, Yang Xiao, “A Survey of Intelligent
training the malicious URL detection system and evaluating its Manuel Sánchez-Paniagua, Eduardo Fidalgo
 Fernández, “Phishing URL Detection: A R Real Case
Scenario Through Login URLs”, IEEE, 2022
2. Detection Designs of HTML URL Phishing Attacks”,
IEEE, 2023
3. Saleem Raja A, Madhubala R, “Survey on Malicious
URL Detection Techniques”, IEEE, 2022
4. Malak Aljabri, Hanan S. Altamimi, “Detecting
Malicious URLs Using Machine Learning Techniques
Review and Research Directions”, IEEE, 2022
5. Vedanti Dantwala, Rishi Lakhani, “A Novel Technique
to Detect URL Phishing based on Feature Count”,
IEEE, 2023
6. Areti Nagendra Soma Charan, Yu-Hung Chen,
“Phishing Websites Detection using Machine
Learning with URL Analysis”, IEEE, 2022
7. Kiruthi Whasan W R, Khiran Khumhaar W R, “Web
Extension for Phishing URL Identification”, IEEE,
2022
8. Rashmi Jha, Gaurav Kunwar, “Machine Learning based
URL Analysis for Phishing Detection”, IEEE, 2023
9. Prakash Pathak; Akhilesh Kumar Shrivas,”
Classification o Phishing Website Using Machine
Learning Based Proposed Ensemble Model”, IEEE,
2023