BACKD00R
An intelligent decision-support system using Random Forest MoE and Weighted Multi-Criteria Analysis for Java projects.
Core Research Features
Built on rigorous academic research to provide intelligent, actionable refactoring recommendations.
Intelligent Detection
Identifies God Class, Feature Envy, and Long Method smells using expert-inspired rule signals.
Data-Driven Prioritization
Ranks refactoring candidates using Weighted Multi-Criteria Analysis (WMCA) considering structural complexity and historical risk.
Safe Sequencing
Implements Deterministic Context-Aware Sequencing (DCAS) to minimize ripple effects and ensure a coherent refactoring order.
Technical Specifications
BACKD00R leverages a sophisticated Mixture of Experts (MoE) architecture combined with Random Forest confidence estimation to provide accurate and reliable code smell detection and refactoring prioritization.
Architecture
Mixture of Experts (MoE)
ML Model
Random Forest Classifier
Confidence
Probability-Based Estimation
Processing
Real-Time Analysis
MoE Architecture Flow
Resource Center
Everything you need to get started with BACKD00R.
Prerequisites
Before installing BACKD00R, ensure your development environment meets these requirements:
- Java 8 or higher installed
- IntelliJ IDEA 2021.1 or later
- Minimum 4GB RAM recommended
Downloads
Get the BACKD00R plugin and the latest version of IntelliJ IDEA to start refactoring smarter.