Cyber Bullying Detection System

A comprehensive system designed to detect and mitigate cyberbullying across multiple media formats, including text, images, and audio. The project leverages machine learning models to identify harmful content and provide a safer environment for users.

🚀 The Stack

Component	Technology
Backend	Python / Flask
Mobile App	Android (Java)
Database	MySQL
Machine Learning	TensorFlow, Keras, Scikit-learn, NLTK
Audio Processing	SpeechRecognition, PyAudio
Networking	Volley (Mobile), demjson3 (API)
Data Handling	Pandas, NumPy

🛠️ Setup Instructions

Follow these steps to get the project running locally:

1. Prerequisites

• Python 3.x installed.
• MySQL Server installed and running.

2. Install Dependencies

Navigate to the Cyber directory and install the required Python packages:

pip install -r requirements.txt

Note

If you encounter issues installing pyaudio on Windows, you may need to download the appropriate .whl file for your Python version from a trusted source or install through pipwin.

3. Database Configuration

1. Open Cyber/database.py and Cyber/setup_db.py.
2. Update the MySQL credentials (user, password, host) to match your local setup.
3. Run the setup script to create the database and tables:

python setup_db.py

This script will automatically create the pythoncyber_bullying1 database and import the schema from RAMANI.sql.

4. Running the Application

Start the Flask server by running:

python main.py

The application will be accessible at http://localhost:5789/.

🔌 API Connection

The system provides a RESTful API for integration. The base URL for all API endpoints is: http://localhost:5789/api

Core Endpoints

Endpoint	Method	Description
/login	GET/POST	User authentication.
/user_upload_regi_file	POST	User registration with profile image.
/user_upload_file	POST	Create a post (Text/Image). Triggers bullying detection.
/upload_audio	POST	Create an audio post. Uses speech-to-text for detection.
/user_comment_post	GET/POST	Add a comment. Scanned for bullying content.
/view_friends	GET/POST	Retrieve list of accepted friends.

📱 Mobile Application (CyberAid)

The project includes a dedicated Android client located in the ImageTextSurgery2 directory. This app allows users to interact with the Cyber Bullying Detection System on the go.

Mobile Tech Stack

• Language: Java
• Platform: Android SDK
• Build System: Gradle
• Communication: Volley for REST API requests.
• Features: Camera integration, Location services, Internal storage access.

Mobile Setup

1. Open the ImageTextSurgery2 folder in Android Studio.
2. Build the project and run it on an emulator or a physical Android device.
3. Upon first launch, use the IP Setting screen to configure the server address (e.g., http://192.168.x.x:5789/).

---

✨ Key Features

• Multi-Modal Detection: Detects bullying in text (Random Forest), images (CNN), and audio (Speech Recognition + NLP).
• Mobile Client: Full-featured Android app for real-time interaction and reporting.
• User Management: Secure login and registration with profile management.
• Social Interaction: Friend requests, post sharing, and commenting system.
• Moderation Tools: Reporting mechanisms and automated content blocking.