AI-powered debate solver

Inspiration

Inspired by Dung's Abstract Argumentation Framework (a foundational theory in AI )

What it does

AI-Powered Debate Solver is a full-stack web application that analyzes complex arguments and determines which positions are logically defensible

How we built it

Backend: Django + Django REST Framework

dung_af.py: Core argumentation framework with weighted semantics graph_optimizer.py: SCC decomposition (Tarjan's), pruning, incremental tracking operators.py: Logic operators for multi-type attacks

Frontend: Vanilla JavaScript + vis.js for graph visualization

Interactive graph with click-to-highlight extensions Responsive UI with optimization statistics

Graph Library:NetworkX for efficient graph operations

Challenges we ran into

For n arguments, we need to check 2^n possible subsets. With 20 arguments, that's 1 million subsets How do we handle negation attacks

Accomplishments that we're proud of

Algorithm Innovation Performance Optimization

What we learned

learnt Django and the algorithm with enhanced features

What's next for AI-powered debate solver

Multi-User Debate Platform Machine Learning Integration

Built With

• 3.1
• 4.2
• css3
• django
• git
• html5
• javascript
• networkx
• pip
• python
• pythonvenv
• vscode