Explainable Plagiarism Analysis Platform

A web-based system project for analyzing plagiarism in academic documents.

Tech Stack

• Frontend: React + Vite + Tailwind CSS
• Backend: FastAPI
• Database: SQLite (later)
• Algorithms: Classical NLP and similarity methods

Project Goals

• Compare one document against another
• Compare one document against a corpus
• Batch compare multiple submissions
• Visualize a class-wide similarity graph
• Detect style shifts inside a document
• Export reports

---

Project Structure

plagiarism-analysis-platform/
│
├── backend/
│   ├── app/
│   ├── uploads/
│   ├── requirements.txt
│   └── .gitignore
│
├── frontend/
│   ├── src/
│   ├── package.json
│   └── .gitignore
│
├── docs/
├── samples/
├── .gitignore
└── README.md

Setup

Prerequisites

Make sure these are installed on your computer:

• Git
• Python 3.11+
• Node.js
• npm

1. Clone the repository

git clone https://github.com/enamee/SDP_Plagiarism_Analysis_Platform.git
cd SDP_Plagiarism_Analysis_Platform

2. Set up the backend

Go to the backend folder:

cd backend

Create a virtual environment.

On Linux/macOS:

python -m venv .venv
source .venv/bin/activate

On Windows Command Prompt:

python -m venv .venv
.venv\Scripts\activate

Install backend dependencies:

pip install -r requirements.txt

3. Set up the frontend

Open a new terminal and go to the frontend folder:

cd frontend
npm install

This installs all frontend packages from package.json.

How to Run the Project

Run the backend

Open Terminal 1.

On Linux/macOS:

cd backend
source .venv/bin/activate
fastapi dev app/main.py

On Windows Command Prompt:

cd backend
.venv\Scripts\activate
fastapi dev app/main.py

Backend will run at:

• http://127.0.0.1:8000
• Health check: http://127.0.0.1:8000/health
• API docs: http://127.0.0.1:8000/docs

Run the frontend

Open Terminal 2:

cd frontend
npm run dev

Frontend will run at:

• http://localhost:5173

Documentation

Additional project notes are available in the docs/ folder:

• docs/demo_script.md — suggested demo flow for presentation
• docs/viva_quick_notes.md — quick viva preparation notes And add this section too:

Known Limitations

• Scanned/image-based PDFs may not extract text correctly
• PDF table reconstruction is less reliable than DOCX
• Sentence splitting is simple and may not be perfect
• Style-shift detection is heuristic/statistical, not full authorship verification
• Graph layout is a simple circular layout for clarity and simplicity