TruthLens - AI Powered News Credibility Analyzer

TruthLens is an AI-powered web application that evaluates the credibility of news articles using a hybrid Machine Learning and Deep Learning pipeline. The system combines a fine-tuned DistilBERT model with a Logistic Regression + TF-IDF classifier to generate a credibility score (0–100). It also integrates the Google Fact Check API to provide additional verification, helping users assess the reliability of online news content.

---

Features

• 📰 AI-powered news credibility analysis
• 🤖 Fine-tuned DistilBERT for contextual understanding
• 📊 Logistic Regression + TF-IDF based classification
• 🌐 Google Fact Check API integration
• 📈 Credibility score ranging from 0–100
• 👤 User authentication system
• 🔐 Secure OTP-based password reset
• 📩 Contact Us form with database storage
• 📱 Responsive and user-friendly interface

---

Tech Stack

Frontend

• HTML
• CSS
• JavaScript

Backend

• Flask
• SQLAlchemy
• Flask-Login
• Flask-Mail

Machine Learning & Deep Learning

• Fine-tuned DistilBERT
• Logistic Regression
• TF-IDF Vectorizer
• spaCy
• NumPy
• Pandas
• Scikit-learn
• Transformers
• PyTorch

Database

• SQLite

---

DistilBERT Training Notebook

The complete Google Colab notebook used for training and fine-tuning the DistilBERT model is available here:

Google Colab: https://colab.research.google.com/drive/1a-TEW8PKknDpJ_ufnkmaxqSHCOFASj74?usp=sharing

---

Fine-tuned DistilBERT Model

Download the fine-tuned DistilBERT model from the link below:

Model: https://drive.google.com/file/d/1-NgDal2jM3q-9vJWl86AZwZnmNcrdY-H/view?usp=sharing

---

Installation

1. Clone the repository

git clone https://github.com/Anisha32414/TruthLens.git
cd TruthLens

2. Create a virtual environment

Windows

python -m venv .venv
.\.venv\Scripts\activate

macOS/Linux

python3 -m venv .venv
source .venv/bin/activate

3. Install dependencies

pip install -r requirements.txt
python -m spacy download en_core_web_sm

4. Configure Environment Variables

Create a .env file in the project root and add the required environment variables:

SECRET_KEY=
DATABASE_URL=
EMAIL=
APP_PASSWORD=
FACT_CHECK_API_KEY=
ADMIN_EMAIL=

5. Run the application

python app.py

Open your browser and visit:

http://127.0.0.1:5000

---

⭐ If you found this project useful, consider giving this repository a Star. Your support is greatly appreciated!