Getting Started

Prerequisites

Installation

git clone https://github.com/rasyidf/Analogi.git
cd Analogi
dotnet build

Running the Desktop App

dotnet run --project Analogi.App

The app opens with a navigation sidebar:

  • Scan — select a folder and run analysis
  • Results — view similarity scores, filter by severity
  • Compare — side-by-side code view with highlighted differences
  • About — version and technology info

Quick Scan (Desktop)

  1. Click Browse and select the folder containing student files
  2. Toggle Submission mode if each student has their own subfolder
  3. Click Start Scan
  4. Review results sorted by similarity percentage
  5. Click any pair to see the side-by-side comparison

Quick Scan (CLI)

# File mode — compare individual files
dotnet run --project Analogi.CLI -- path/to/files --top 10

# Submission mode — compare student folders
dotnet run --project Analogi.CLI -- path/to/submissions --submissions --top 10

Understanding Results

Each pair receives a Similarity Index (0–100%) based on a weighted average of multiple analyzer scores:

Level Threshold Interpretation
Extreme ≥ 90% Almost certainly copied
Very High ≥ 80% Very likely copied with minor changes
High ≥ 70% Significant structural similarity
Moderate ≥ 60% Suspicious, worth investigating
Low ≥ 40% Some shared patterns
Minor ≥ 20% Minimal similarity (likely coincidence)
Original < 20% Independent work

A high similarity score doesn’t automatically mean plagiarism. Common patterns (boilerplate, standard algorithms) can produce false positives. Always review the specific reasons listed for each pair.