MSMS - Mentor-based Student Management System

MSMS is a comprehensive platform designed for mentors to professionally manage students preparing for the university entrance exam (YKS). The system digitizes the coaching process and includes a unique AI-powered book analysis engine that predicts book popularity based on sales data.

About the Project

This platform bridges the gap between mentors and students through data-driven tracking. It operates on a Service-Oriented Architecture (SOA), integrating multiple technologies (Node.js, Python/ML, gRPC) to provide a seamless experience.

Core Feature: AI Book Analysis Module

The system utilizes a Machine Learning regression model to help students make informed decisions about study materials.

• The Logic: The model analyzes technical attributes of a book to predict its sales volume.
• The Goal: Predicted sales figures serve as a proxy for "popularity." This helps students prioritize resources that are widely used and trusted by the community, optimizing their study time.

Tech Stack

The project is built using a modern stack and microservice-like communication patterns.

Area	Technologies
Frontend	React, Vite, Tailwind CSS, Context API
Backend	Node.js, Express.js, JWT Auth
Database	SQL (PostgreSQL), Raw SQL/Sequelize
AI / ML	Python, Scikit-Learn, Pandas, gRPC (Protobuf)
Identity	Mock SOAP Web Service (WSDL)
Communication	REST API, gRPC, SOAP

Architecture & Directory Structure

The system consists of four main distinct components:

1. backend/: The main application server. It handles client requests, manages the database, and communicates with the AI service via gRPC.
2. frontend/client/: The React-based user interface for Mentors, Students, and Admins.
3. external-services/PythonMLEngine/: A Python-based gRPC server that hosts the ML model for book sales predictions.
4. external-services/MockIdentityProvider/: A SOAP service simulating a government ID (TC Identity) verification system.

Key Features

Mentor Panel

• Student Management: Add students, verify identities via SOAP, and manage profiles.
• Academic Tracking: Input trial exam results and visualize progress charts.
• Scheduling: Organize meetings and study sessions (Calendar integration).
• Task System: Assign specific tasks to students and track completion status.

Student Panel

• Performance Analytics: View exam results and personal growth graphs.
• AI Book Assistant: Query books to see popularity predictions and AI-driven insights.
• My Schedule: Track weekly lessons and mentor meetings.

Installation & Setup

Follow these steps to run the project locally.

1. Database Setup

Initialize the database using the scripts found in the backend/database/ folder.

• Run create-db.js or the relevant .sql script to create tables.
• Run seed-db.js to populate the database with initial dummy data.

2. Backend (Node.js) Setup

cd backend
npm install
# Ensure you configure your .env file with DB credentials
node server.js

3. AI Engine (Python) Setup

For the book recommendation feature to work, the gRPC server must be running:

cd external-services/PythonMLEngine
pip install -r requirements.txt
python grpc_server.py

4. Mock Identity Provider (Optional)

To simulate the ID verification service:

cd external-services/MockIdentityProvider
npm install
node server.js

5. Frontend (React) Setup

cd frontend/client
npm install
npm run dev

AI Model Training

The model training process is documented in external-services/BookSalesPrediction. The system uses a Gradient Boosting Regressor model trained on the Kitapyurdu dataset (located in datasetVersions), which yielded the highest accuracy metrics among tested algorithms.

Documentation

For detailed technical documentation, please refer to the docs/ folder:

• SRS: Software Requirement Specification
• DDD: Database Design Document
• Architecture: SOA Diagrams and Data Flow

License: MIT