This README contains all the necessary info about the project. Scroll down to the "Getting Started" section to see how to add it to your Chrome browser.

Watch a video demonstration of the project: Click Here

Project Darpan (दर्पण)

Reflecting the Truth in Digital Discourse

---

📖 Overview

Darpan (दर्पण), the Sanskrit word for "Mirror," is a powerful analysis engine designed to reflect the true nature of online content. In an era where digital discourse is often clouded by toxicity, misinformation, and hate, Project Darpan serves as a mirror, exposing harmful patterns that undermine constructive conversation.

This tool provides an objective, pattern-based analysis to identify and score various forms of negative content, from subtle propaganda tactics to overt hate speech. By holding a mirror to the text, we aim to empower developers, moderators, and researchers to build healthier and safer online communities.

---

🎯 The Problem

The digital public square is increasingly polluted with:

• Separatist and Anti-National Rhetoric that threatens national integrity.
• Hate Speech and Communal Incitement that divides communities.
• Geopolitical Disinformation that spreads false narratives.
• Attacks on Institutions that erode public trust.
• Subtle Propaganda Tactics that manipulate public opinion.

Moderating this content manually is unsustainable. Project Darpan offers an automated, scalable solution to identify and categorize these toxic patterns effectively.

---

✨ Key Features

• Toxicity Scoring: Each piece of text is assigned a cumulative "Toxicity Score" based on the severity and frequency of detected patterns.
• Detailed Categorization: Content is analyzed and flagged under specific categories of toxicity for granular insight.
• Pattern-Based Detection: Utilizes a comprehensive and expanding library of keywords, regular expressions, and linguistic patterns.
• Context-Aware: The patterns are curated with a focus on the socio-political context of the Indian digital landscape.
• Extensible Framework: The toxicPatterns object can be easily updated to include new keywords, patterns, and categories as online discourse evolves.

---

🔬 How It Works

Project Darpan operates on a weighted pattern-matching system. The core logic iterates through a predefined set of toxic patterns:

1. Ingestion: The tool takes raw text as input.
2. Keyword Analysis: It scans the text for keywords associated with various toxic categories (e.g., separatist, hate_speech, military_disinfo).
3. Regex Pattern Matching: It applies regular expressions to identify manipulative tactics, such as excessive capitalization, whataboutism, or conspiracy theory terminology.
4. Scoring & Aggregation: When a pattern is matched, its corresponding score is added to a running total. The system also logs which specific patterns were triggered.
5. Output: The final output is a JSON object containing the total toxicity score and a breakdown of all the detected issues and their categories.

---

🗂️ Toxicity Categories

The engine currently identifies and scores content across several major categories, including:

High Severity:

• Separatist & Anti-National Rhetoric
• Casteist Abuse
• Misogyny & Gender-Based Abuse
• Hate Speech & Communal Incitement

Medium Severity:

• Military & Security Forces Disinformation
• Geopolitical Disinformation
• Institutional Attacks & Distrust

Low Severity:

• General Negativity & Demoralization
• Propaganda Tactics & Manipulation

---

🚀 Getting Started

You can use Project Darpan either as a Node.js module in your own projects or directly in your browser as a Chrome Extension.

As a Chrome Extension (For Direct Use)

Follow these steps to load the extension into your Chrome browser:

1. Download the Project:

   • Clone the repository to your local machine:

     git clone https://github.com/your-username/project-darpan.git

   • Alternatively, you can download the project as a ZIP file and extract it.
   • Generate your Hugging Face API key and replace it with (apiKey: "your-huggingface-api-key-here") in popup.js.

2. Open Chrome Extensions Page:

   • Launch Google Chrome.
   • Navigate to chrome://extensions in your address bar.

3. Enable Developer Mode:

   • In the top-right corner of the Extensions page, find and turn on the Developer mode toggle. This will reveal a new menu bar.

4. Load the Extension:

   • Click the "Load unpacked" button that appears on the top-left.
   • A file selection window will open. Navigate to and select the entire project-darpan folder that you cloned or unzipped.
   • Click "Select Folder".

5. Ready to Go!:

   • Project Darpan will now appear in your list of extensions. Ensure it's enabled with the toggle switch. The extension is now active and will analyze content as you browse.

---

🤝 Contributing

Contributions are welcome! If you wish to improve Project Darpan, please feel free to fork the repository, make your changes, and submit a pull request. You can contribute by:

• Adding new keywords to existing categories.
• Creating new categories of toxic behavior.
• Improving the regular expressions for pattern matching.
• Enhancing the scoring algorithm.