Skip to content

This repository contains an AI-Based Threat Intelligence and Prediction System to detect and classify phishing URLs using machine learning. It includes data preprocessing, model training, and evaluation with Logistic Regression and Multinomial Naive Bayes, aiming to enhance cybersecurity by identifying malicious URLs.

Notifications You must be signed in to change notification settings

Tech-Anshika/AI-Based-Threat-Intelligence-and-Prediction-System

Repository files navigation

AI-Based-Threat-Intelligence-and-Prediction-System

This repository contains an AI-Based Threat Intelligence and Prediction System to detect and classify phishing URLs using machine learning. It includes data preprocessing, model training, and evaluation with Logistic Regression and Multinomial Naive Bayes, aiming to enhance cybersecurity by identifying malicious URLs.

🔍 Project Overview:

Phishing attacks pose a significant threat to online security, and traditional methods often fall short in identifying these rapidly evolving threats. To address this, I implemented a machine learning solution leveraging Logistic Regression and Multinomial Naive Bayes algorithms to classify URLs as malicious or benign.

💡 Key Features:

  • Data Preprocessing: Tokenization and stemming of URL text data to prepare it for model training.
  • Model Training: Leveraging state-of-the-art algorithms for accurate phishing URL classification.
  • Model Evaluation: Detailed performance metrics, including classification reports and confusion matrices.
  • Visualization: Interactive charts and graphs to visualize model performance.
  • Serialization: Efficient saving and loading of trained models for future use.

📊 Results:

The system demonstrated impressive accuracy in detecting phishing URLs, showcasing the potential of AI to significantly improve cybersecurity measures.

About

This repository contains an AI-Based Threat Intelligence and Prediction System to detect and classify phishing URLs using machine learning. It includes data preprocessing, model training, and evaluation with Logistic Regression and Multinomial Naive Bayes, aiming to enhance cybersecurity by identifying malicious URLs.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published