IntelliJ IDEA Plugin

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

Java Source Code
God Class Expert
Feature Envy Expert
Long Method Expert
Random Forest Gating Network
Prioritized Refactoring Queue

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.