EchoGraph - Social Echo Chamber Analyzer

A comprehensive tool for analyzing social media echo chambers and misinformation pathways using knowledge graphs.

🎯 Project Overview

EchoGraph analyzes social media data (Reddit/Twitter) to:

• Build knowledge graphs from user interactions
• Detect echo chambers and filter bubbles
• Identify misinformation pathways
• Visualize community structures
• Provide early warning for coordinated misinformation campaigns

Demo:

https://drive.google.com/file/d/12CCHeLH6UgBxAytda2WyfnTBSl2OUAdK/view?usp=sharing

🏗️ Architecture

echoGraph/
├── src/
│   ├── knowledge_graph/     # Custom KG implementation
│   ├── data_collection/     # Reddit/Twitter scrapers
│   ├── analysis/           # Echo chamber & misinformation detection
│   ├── visualization/      # Graph visualization
│   └── web_interface/      # Flask web app
├── data/                   # Sample datasets
├── models/                 # Trained models
├── tests/                  # Unit tests
└── notebooks/              # Jupyter analysis notebooks

🚀 Features

Core Knowledge Graph

• Custom graph implementation with nodes and edges
• Support for user, post, and content entities
• Relationship modeling (follows, shares, comments)
• Graph traversal and analysis algorithms

Echo Chamber Detection

• Community detection using custom algorithms
• Echo chamber scoring based on content diversity
• Polarization measurement
• Filter bubble identification

Misinformation Analysis

• Content similarity analysis
• Coordinated behavior detection
• Information pathway tracing
• Early warning system

Visualization

• Interactive network graphs
• Community highlighting
• Temporal analysis views
• Dashboard with metrics

📊 Datasets

The project works with:

• Reddit post and comment data
• User interaction networks
• News article URLs and metadata
• Optional fact-checking labels

🛠️ Installation

# Clone the repository
git clone <https://github.com/HawksLab/echoGraph>
cd echoGraph

# Create virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Run the application
python src/main.py

🎮 Usage

1. Data Collection

from src.data_collection.reddit_scraper import RedditScraper

scraper = RedditScraper()
data = scraper.collect_subreddit_data(['politics', 'conspiracy'], days=30)

2. Build Knowledge Graph

from src.knowledge_graph.graph import KnowledgeGraph

kg = KnowledgeGraph()
kg.build_from_reddit_data(data)

3. Analyze Echo Chambers

from src.analysis.echo_chamber_detector import EchoChamberDetector

detector = EchoChamberDetector(kg)
chambers = detector.detect_echo_chambers()
scores = detector.calculate_echo_scores()

4. Visualize Results

from src.visualization.graph_visualizer import GraphVisualizer

viz = GraphVisualizer(kg)
viz.create_interactive_plot(chambers)
viz.save_html("echo_chambers.html")

📈 Analysis Methods

Echo Chamber Detection

• Modularity Analysis: Identifies tightly connected communities
• Content Diversity Score: Measures information variety within groups
• Cross-Group Interaction: Analyzes inter-community connections
• Temporal Dynamics: Tracks chamber formation over time

Misinformation Detection

• Content Similarity: Identifies repeated/coordinated content
• Propagation Patterns: Traces information spread pathways
• User Behavior Analysis: Detects suspicious coordination
• Network Anomalies: Identifies unusual connection patterns

🧪 Algorithms Implemented

1. Custom Graph Algorithms

   • Breadth-First Search (BFS)
   • Depth-First Search (DFS)
   • Shortest Path (Dijkstra's)
   • Community Detection (Louvain-inspired)

2. Echo Chamber Metrics

   • Polarization Index
   • Diversity Score
   • Isolation Coefficient
   • Homophily Measure

3. Misinformation Detection

   • Content Similarity (TF-IDF + Cosine)
   • Behavioral Anomaly Detection
   • Cascade Analysis
   • Coordination Scoring

📊 Sample Results

The system provides:

• Interactive network visualizations
• Echo chamber reports with scores
• Misinformation pathway maps
• Temporal analysis charts
• User and content analytics

🔬 Research Applications

This project can be used for:

• Academic research on social media dynamics
• Digital humanities studies
• Computational social science
• Media literacy education
• Platform policy research