The research introduces an innovative ML-driven system designed to combat phishing attacks and malicious emails. It uniquely adopts a comprehensive approach, encompassing dataset collection, preprocessing, ML model development, and a user-friendly browser extension interface. Achieving so by facilitating the process of identifying phishing URLs through XGBoost and detecting malicious email content using Multinomial Naive Bayes, it exemplifies heightened cybersecurity protocols. The dataset mirrors diverse phishing scenarios, guided by thorough Exploratory Data Analysis (EDA). The XGBoost model utilizes URL features for precise phishing website identification, while the Multinomial Naive Bayes excels in discerning malicious email content. Implemented as a browser extension, the graphical user interface ensures seamless integration into users' digital landscapes, enabling immediate feedback on encountered threats for swift mitigation and user awareness.
The study presents a machine learning-based system to mitigate malicious emails and phishing attempts. It utilizes a thorough methodology that includes gathering datasets, preprocessing, creating ML models, and creating an intuitive browser extension interface. The system detects malicious email content with 98.6% accuracy and phishing URLs with 96.4% accuracy. This same seamless integration of the graphical user interface into users' digital environments is implied.
cd frontend
npm run build
""" Go to chrome, go to extensions , turn on developers mode , select 'Load Unpacked' , select the build folder so created.
You'll have the extension added in Chrome. """THE BACKEND IN DEPLOYED AND INTEGERATED
BUT IF YOU WANT TO RUN THE BACKEND LOCALLY
# in another terminal
cd backend
pip install -r requirements.txt
python app.py
#(make sure to change the routes in the frontend)
