🕵️‍♂️ Plagiarism Detection Engine

This project is a C++-based plagiarism detection tool designed to identify similar code submissions using efficient algorithms like tokenization, hashing, winnowing (fingerprinting), and Jaccard similarity.

---

🚀 Features

• 📁 Detect plagiarism across entire folders of code submissions
• 🧠 Uses n-gram hashing and winnowing for efficient fingerprint generation
• 📊 Calculates Jaccard similarity between pairs of submissions
• 👨‍🎓 Supports filtering by problems and users

---

🛠️ How It Works

1. Preprocessing

   • Removes both single-line (//) and multi-line (/* ... */) comments
   • Trims trailing whitespaces
   • Tokenizes the source code

2. Tokenization & Hashing

   • Splits code into n-grams (default n=3)
   • Applies polynomial rolling hash:

     hash = Σ (char[i] - '!') × PRIME^i
     

3. Winnowing (Fingerprinting)

   • Slides a window of size WINDOW_SIZE (default 5)
   • Selects the minimum hash from each window as a stable fingerprint

4. Similarity Comparison

   • Computes Jaccard similarity:

     Jaccard(A, B) = |A ∩ B| / |A ∪ B|
     

   • Compares fingerprints of code pairs to detect similarity

---