Anti Phishing Academy

Our project consists of a web application which enables companies to train their employees to differentiate between ham emails and scam emails and provides the employer with their respective results.

Comment

Inspiration

Our project is inspired by the SB Software Challenge, which focuses on implementing AI in Cybersecurity. We wanted to combine AI with an educational aspect to help employees recognize spam emails (phishing), which are common and pose significant risks in modern society.

What It Does

Anti-Phishing Academy is a web application that:

  • Uses a logistic regression ML model trained on over 10,000 records to predict whether an email is spam with 94% accuracy.
  • Allows employees to create an account and take quizzes on identifying spam emails.
  • Draws quiz samples from a pool of ~5,000 emails.
  • Provides admin dashboards to track employee participation and quiz results.

How It Was Built

Machine Learning Model:

  • Language: Python
  • Libraries:
    • scikit-learn – training the logistic regression model
    • joblib – model serialization
    • numpy & pandas – data processing and manipulation
    • matplotlib & seaborn – data visualization
  • Datasets: Two datasets totaling over 10,000 records

Web Application:

  • Frontend: HTML, CSS, JavaScript
  • Backend: PHP – user authentication and quiz management
  • Integration: Flask (Python) APIs – connecting ML model with the web app

Challenges

  • Integrating ML with the Web App: Ensuring smooth communication between the Python ML model and PHP web application.
  • Data Handling: Cleaning and preparing datasets for accurate model training.
  • Real-Time Quiz Feedback: Ensuring the web application could interact with the ML model efficiently during quizzes.

Accomplishments

  • Delivered a fully working, user-friendly web app in a short time frame.
  • Successfully trained a logistic regression ML model from scratch.
  • Seamlessly integrated two separate processes (Python ML model + PHP web app).

What We Learned

  • How to integrate machine learning into a web application.
  • Techniques for connecting different processes and ensuring smooth communication.
  • Importance of high-quality training data for ML accuracy.
  • Project planning and collaboration under a hackathon time constraint.

What's Next For Anti-Phishing Academy

  • Increase the size of the training dataset to improve ML model accuracy.
  • Expand the web app to include more advanced phishing detection features.
  • Potentially create real-time email scanning for enterprise users.

Deploying the Project

  • Run the "submission_checker.py" file in the "BackEnd" folder.
  • Locally host the website found in the "FrontEnd" folder.
  • Initialize database parameters properly in "db_connection.php".
Share this project:

Updates