The homepage - dark mode
The analysis generated from an article - dark mode
The article with sentiment highlighted - dark mode
The homepage - light mode
The analysis generated from an article - light mode
The article with sentiment highlighted - light mode

The Credibility Compass

A news credibility analysis tool developed for Bath Hack 2025 that evaluates news articles for factual accuracy, political bias, source reliability, and sentiment.

🌟 Inspiration

In an era of information overload and increasing concerns about misinformation, we were inspired to create a tool that helps people navigate the complex media landscape. The Credibility Compass was born from our desire to empower readers with AI-assisted insights into the credibility and bias of the news they consume.

Our goal was to build a solution that:

Promotes media literacy and critical thinking
Provides transparent assessments of news content
Helps users recognize political bias and sentiment in reporting
Reduces the spread of misinformation by highlighting factual inaccuracies

🔍 What It Does

The Credibility Compass allows users to:

Enter any news article URL for comprehensive analysis
View factuality assessments with confidence scores, ratings, and supporting sources
Understand the reliability of the news source with detailed reasoning
Identify political leaning of both the article and its source on a spectrum from Far Left to Far Right
Analyze sentiment patterns throughout the article, with entity-specific sentiment highlighting

🛠️ How We Built It

This project is built with a modern web stack:

Frontend: Next.js with the App Router, Tailwind CSS, and shadcn/ui components
Backend: Next.js API routes handling content processing and analysis
AI Integration:
- Google's Gemini API for comprehensive content analysis
- Perplexity API for search-based factual verification and source credibility assessment
Data Processing: Custom algorithms for entity extraction, sentiment analysis, and bias detection

🧠 Challenges We Faced

Building The Credibility Compass came with several challenges:

Content Extraction: Developing robust methods to extract clean article text from diverse news websites
Bias Detection: Creating a nuanced system for identifying political bias without introducing our own biases
Performance Optimization: Balancing comprehensive analysis with reasonable response times
Source Credibility: Establishing reliable metrics to evaluate the trustworthiness of news sources
UI/UX Design: Creating an intuitive interface that presents complex information clearly

🏆 Accomplishments

We're proud of creating:

A fully functional tool that provides multi-faceted analysis of news content
An intuitive user interface that makes complex information accessible
A system that balances detailed analysis with user-friendly presentation
Integration with advanced AI models to provide sophisticated insights

📚 What We Learned

Through this project, we gained expertise in:

Prompt engineering for specialized AI analysis
Content extraction techniques for web articles
Political bias detection methodologies
Sentiment analysis implementation
Building responsive UIs for data-heavy applications
Optimizing API calls to external AI services

🔮 What's Next

Future enhancements we're considering:

Browser extension for instant analysis while browsing
Expanded historical context for news sources
Comparison feature to analyze multiple articles on the same topic
Community feedback integration to improve analysis accuracy
Mobile application development

🚀 Getting Started

Installation

# Clone the repository
git clone https://github.com/CDE90/BathHack2025.git
cd BathHack2025

# Install dependencies
pnpm install

Development

# Start the development server
pnpm dev

Building for Production

# Build the application
pnpm build

# Start the production server
pnpm start

🧪 Testing and Quality Assurance

# Run linting
pnpm lint

# Fix linting issues
pnpm lint:fix

# Run type checking
pnpm typecheck

# Check formatting
pnpm format:check

# Fix formatting
pnpm format:write