Information Noise Detector

Gain Leverage on Information. Find Your Pivot Point in a World of Noise.

A browser extension that analyzes web pages for emotional manipulation, information noise, and credibility signals - helping you navigate the modern information landscape with clarity and confidence.

🚀 Quick Start

Installation

Chrome/Edge/Chromium-based Browsers

1. Download the Extension

   • Clone or download this repository to your computer
   • Extract the files to a folder (e.g., fulcrum-extension)

2. Enable Developer Mode

   • Open Chrome/Edge and go to chrome://extensions/ (or edge://extensions/)
   • Toggle "Developer mode" in the top-right corner

3. Load the Extension

   • Click "Load unpacked"
   • Select the folder containing the extension files
   • The extension icon should appear in your browser toolbar

4. Configure Settings (Optional)

   • Click the extension icon and select "Settings"
   • Configure LLM API for enhanced analysis (requires OpenAI-compatible API)
   • Adjust detection sensitivity and other preferences

Firefox

1. Go to about:debugging
2. Click "This Firefox"
3. Click "Load Temporary Add-on"
4. Select the manifest.json file from the extension folder

🎯 How to Use

1. Visit any webpage - the extension automatically analyzes content
2. Click the extension icon to view the noise assessment
3. Toggle highlights to see problematic words marked on the page
4. Expand sections to see detailed analysis and trigger words
5. Find better sources with the built-in alternative search feature

The Problem

• Anxiety as a Business Model: Much of today's online content is engineered to provoke an emotional response—fear, outrage, or FOMO—to drive engagement and sales.
• Information Noise & Cognitive Overload: The sheer volume of low-quality, speculative, and emotionally charged information makes it difficult to identify what is true and important.
• Erosion of Agency: When presented with constant crises beyond our control, we risk falling into a state of learned helplessness, losing focus on the areas where we can make a real difference.

Our Solution: A Tool for Empowerment

The Fulcrum works as a browser extension that analyzes web pages in real-time. It provides a simple, non-intrusive dashboard that reveals the hidden characteristics of the content you are viewing, allowing you to assess its quality and intent with clarity.

We believe that the best defense against bad information is not to hide it, but to understand it. The Fulcrum empowers your judgment; it does not replace it.

📊 Features

Real-Time Analysis Dashboard

• Overall Noise Score (0-100) - instant assessment of information quality
• Emotional Intensity - detects anxiety, outrage, FOMO, and hyperbolic language
• Source Credibility - evaluates domain reputation and journalistic standards
• Actionability Index - measures practical usefulness vs. abstract speculation
• Logical Fallacy Detection - identifies manipulative reasoning patterns

Smart Detection Methods

• 🤖 LLM-Powered Analysis - uses AI to identify nuanced manipulation patterns
• 📝 Pattern Matching - fallback system for reliable basic detection
• 💾 Intelligent Caching - avoids redundant API calls to save costs
• 🎯 Percentage-Based Scoring - accurate assessment regardless of content length

User Experience

• 🖍️ Text Highlighting - marks problematic words with explanatory tooltips
• 📱 Compact Interface - collapsible sections save space
• 🔍 Alternative Source Finder - suggests higher-quality sources
• ⚙️ Customizable Settings - adjust sensitivity and detection parameters

Current Capabilities

✅ Emotional Manipulation Detection

• Anxiety triggers: "urgent", "crisis", "breaking", "emergency"
• Outrage triggers: "scandal", "disgusting", "corrupt", "lies"
• FOMO triggers: "secret", "exclusive", "limited time", "insider"
• Hyperbolic language: "extremely", "absolutely", "never", "always"

✅ Source Analysis

• Domain reputation checking
• Author and date verification
• External link assessment
• Educational/government source recognition

✅ Logical Fallacy Detection

• Ad hominem attacks
• Appeals to emotion
• False dichotomies
• Percentage-based frequency analysis

✅ Actionability Assessment

• Instructional language detection
• Contact information verification
• Abstract vs. concrete language analysis
• Structured content recognition

⚙️ Configuration

LLM API Setup (Enhanced Analysis)

For advanced AI-powered analysis, configure an OpenAI-compatible API:

1. Get API Access

   • OpenAI API: https://platform.openai.com/api-keys
   • Or any OpenAI-compatible endpoint (Anthropic, local models, etc.)

2. Configure in Settings

   • API URL: https://api.openai.com/v1 (or your endpoint)
   • API Key: Your API key
   • Model: gpt-4 (or your preferred model)
   • Custom Prompt: Modify detection criteria

3. Cost Optimization

   • Extension automatically caches analysis results
   • No repeated API calls for the same page until refresh
   • Typical cost: $0.01-0.05 per page analysis

Scoring System

The extension uses percentage-based scoring for accuracy:

• Emotional Intensity: 25% noise words = 50 points, 50% = 100 points
• Actionability: 5% actionable content = 100 points
• Logical Fallacies: 25% fallacy words = 50 points, 50% = 100 points
• Overall Score: Weighted average of all metrics

🔧 Troubleshooting

Common Issues

Extension not loading:

• Ensure all files are in the same folder
• Check that Developer Mode is enabled
• Try refreshing the extensions page

No analysis appearing:

• Check browser console for errors (F12 → Console)
• Verify the page has loaded completely
• Try refreshing the webpage

LLM analysis not working:

• Verify API URL and key in Settings
• Check API credits/quota
• Try switching to pattern matching mode

Highlights not showing:

• Click "Toggle Highlights" button
• Check if analysis has completed
• Some sites may have CSS conflicts

Browser Compatibility

✅ Supported: Chrome, Edge, Brave, Opera, Chromium-based browsers
✅ Firefox: Manual installation required
❌ Safari: Not currently supported (Manifest V3 required)

🔮 Roadmap

coming soon

🤝 Contributing

We welcome contributions to make this tool more effective at detecting information noise:

Ways to Help

• 🐛 Bug Reports: Found an issue? Open an issue with details
• 💡 Feature Requests: Ideas for better noise detection
• 🔧 Code Contributions: Improve algorithms, UI, or add features
• 📊 Testing: Try the extension and report what works/doesn't work
• 📚 Documentation: Help others understand and use the tool

Development Setup

1. Fork the repository
2. Make your changes
3. Test the extension locally
4. Submit a pull request with clear description

Key Files

• content.js - Main analysis engine
• popup.html/js/css - User interface
• background.js - Service worker
• options.html/js - Settings page
• manifest.json - Extension configuration

📄 License

This project is licensed under the AGPLv3 License - feel free to modify and distribute, provided you adhere to the terms of the AGPLv3.

🎯 Mission

We believe the best defense against bad information is not to hide it, but to understand it. This extension empowers your judgment; it does not replace it.

Our goal is to build a calmer, more intentional internet where people can navigate information with confidence and clarity.

---

Built with the conviction that information literacy is a fundamental skill for the digital age.

Last updated: July 24, 2025