Skip to content

NewsGuard is a machine learning system built to detect fake or misleading news articles. By leveraging Natural Language Processing (NLP), it classifies text as Legitimate or Fake News, providing a confidence score and an explanation using LIME.

Notifications You must be signed in to change notification settings

ishratarshad/newsguard

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

37 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

NewsGuard


Project Overview

NewsGuard is a machine learning system built to detect fake or misleading news articles. By leveraging Natural Language Processing (NLP), it classifies text as Legitimate or Fake News, providing a confidence score and an explanation using LIME.


Folder Structure

newsguard/
│
├── datasets/           # Raw and cleaned data files
├── analysis/           # EDA reports and data visualizations
├── docs/               # Documentation, test results, and API usage
├── scripts/            # Python scripts (training, testing, LIME, etc.)
├── model/              # Saved models, vectorizers, and features
├── api/                # Flask API implementation
├── ui/                 # (Optional) React front-end for user interaction
│
├── README.md           # Project overview and instructions
├── requirements.txt    # Required Python libraries
├── .gitignore          # Files/directories to ignore in version control

Features

  • Fake News Classifier using Naive Bayes & TF-IDF
  • REST API with /predict and /explain routes
  • LIME Explainability for interpretable results
  • Confidence score for predictions
  • Clean documentation and test logs
  • Tested with Postman + documented test outputs

Tech Stack

  • Python (NLP & API)
  • Scikit-learn, LIME, Pandas
  • Flask (Backend API)
  • Postman (API testing)

About

NewsGuard is a machine learning system built to detect fake or misleading news articles. By leveraging Natural Language Processing (NLP), it classifies text as Legitimate or Fake News, providing a confidence score and an explanation using LIME.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published