🔗 Twitter Content GNN Edge Prediction

Results & Analysis Portal

Graph Neural Network Analysis of Social Network Links

Welcome to the GNN Edge Prediction Results Portal. This site contains comprehensive analyses and visualizations from our Graph Neural Network experiments on Twitter network data.

Explore the different sections below to find performance matrices, interpretations, and detailed analyses across multiple graph types and feature encodings.

📊 Interpretations & Analysis
Executive Summary START HERE

Quick overview of key findings and recommendations

📈 Full Analysis Report

Comprehensive interpretation of experimental findings, content vs. structure analysis, and dataset context

🌐 Graph Shqip NSL

Primary network analysis with NSL variant

📊 AUC Performance Matrix

AUC-ROC scores for all model and feature combinations

Accuracy Matrix

Accuracy scores across all experiments

🌐 Graph Shqip

Standard network analysis variant

📊 AUC Performance Matrix

AUC-ROC scores for all model and feature combinations

Accuracy Matrix

Accuracy scores across all experiments

🔄 Graph Shqip RT

Retweet network analysis variant

📊 AUC Performance Matrix

AUC-ROC scores for all model and feature combinations

Accuracy Matrix

Accuracy scores across all experiments

📚 Reference
📖 Graph Types Documentation

Detailed descriptions and comparisons of different graph variants

📁 Raw Data
📊 Results CSV

Machine-readable results data in CSV format

📄 Results Text

Human-readable results summary

📋 Quick Navigation Guide

  • New to this analysis? Start with the AI Interpretation Report for context and key findings
  • Want to compare metrics? Check the performance matrices (AUC and Accuracy)
  • Interested in specific graphs? Navigate to NSL, Standard, or RT variants
  • Looking for raw data? Download CSV or TXT files for further analysis