Scribd
Upload
43 views8 pages

BSC Final Project

The document presents a project on AI-based phishing detection tools, focusing on email monitoring and website protection to combat the increasing prevalence of phishing attacks. It discusses the limitations of traditional detection methods and proposes a machine learning model to enhance real-time detection accuracy. The study aims to evaluate existing algorithms and explore future applications of the developed model in real-world scenarios.

Uploaded by

Arani Navaratnarajah
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
43 views8 pages

BSC Final Project

The document presents a project on AI-based phishing detection tools, focusing on email monitoring and website protection to combat the increasing prevalence of phishing attacks. It discusses the limitations of traditional detection methods and proposes a machine learning model to enhance real-time detection accuracy. The study aims to evaluate existing algorithms and explore future applications of the developed model in real-world scenarios.

Uploaded by

Arani Navaratnarajah
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 8

AI Based Phishing Detection

Tools
Project Presentation
By AWM. Saajith
AI-Based Phishing Detection Tools Introduction

Email Monitoring
1
Integrating AI algorithms to continuously monitor email communications
for potential phishing attempts and alert users in real-time, significantly
reducing the risk of data breaches.

Website Protection
2
Employing AI-driven tools to analyze website traffic and detect
fraudulent domains or phishing sites, providing users with immediate
warnings and enhanced protection against cyber threats.
Research Background: Phishing Threat Landscape

Prevalence
1
Phishing attacks are growing rapidly, requiring advanced
measures to counteract the evolving complexity of threats.

Detection Gaps
2
Traditional detection methods are inadequate, creating
opportunities for AI-driven tools to enhance real-time
identification.

AI Solutions
3
Machine learning algorithms can analyze vast data to
effectively recognize patterns and anomalies in phishing
attempts.
Scope, Aim, and Objectives of the Study

Research Focus

Understanding the role of AI in identifying phishing threats effectively.


1

Algorithm Evaluation
2 Assessing existing AI algorithms for their efficiency in phishing
detection tasks.

Identifying Challenges
Recognizing limitations of current tools to improve 3
phishing detection capabilities.

Proposed Model
4
Developing a machine learning model that
increases detection accuracy in real-time
scenarios.
Literature Review: Historical and Current Approaches

Description Year Accuracy Techniques

Historically used to
Rule-Based Systems N/A Varying effectiveness Heuristic analysis
detect phishing

Utilizes advanced
AI-Driven Approaches 2022 Up to 96% Random Forest, SVM
algorithms

Machine Learning Utilizes algorithms like NLP and machine


2022 Up to 96%
Models SVM learning

Challenges & False positives and


N/A N/A Model scalability
Limitations adaptability issues
Study Methodology & Tools

 Utilizes machine learning tools like Python, TensorFlow, and Scikit-learn.

 Data from public phishing datasets pre-processed.

 Uses supervised learning algorithms like Random Forest, SVM, and NLP.

 Uses deep learning methods like NLP.

 Measures model performance using confusion matrix, precision, memory, and F1-score.
Conclusion & Future Work

To make cybersecurity stronger in today's digital world, we need phishing


detection tools that are built on AI. The goal of this study is to improve the
accuracy of phishing spotting by creating a machine learning model that
can spot phishing in real time. In the future, researchers will look into
ways to use this model in real life and investigate unsupervised learning
for dynamic phishing risks.
Thank You

You might also like