ML-Dopamind 🧠💬

Project Overview

ML-Dopamind is a machine learning project focused on emotion detection using Natural Language Processing (NLP). The project aims to classify emotional states in text conversations.

Emotion Classification Categories

• Happiness
• Sadness
• Fear
• Anger

Repository Structure

Branches

• main: Primary development branch
• verry: Individual development branch
• richal: Individual development branch
• kiki: Individual development branch

Technical Specifications

Model Architecture

• Framework: TensorFlow
• Pretrained Model: IndoBERT
• Tokenizer: BertTokenizer
• Classification: 4-class emotion detection

Key Technologies

• Python
• TensorFlow
• Transformers (Hugging Face)
• Flask
• JWT Authentication

Key Features

• Real-time emotion detection
• Multilingual support (Indonesian)
• Secure API endpoint
• Logging and error handling

API Endpoint

/predict

• Method: POST
• Authentication: Bearer Token
• Input: Text or list of texts
• Output: Emotion prediction

Installation

Prerequisites

• Python 3.8+
• TensorFlow
• Transformers
• Flask
• JWT

Setup Steps

1. Clone the repository
2. Install dependencies
3. Download pre-trained weights
4. Configure environment variables

Usage Example

# Sample prediction request
{
    "texts": ["Aku sedang merasa senang hari ini"],
    "Authorization": "Bearer <your_token>"
}

Security Features

• JWT Token Authentication
• Input validation
• Error logging
• Secure model inference

Deployment

• Gunicorn server
• Configurable port
• Cloud-ready configuration

Development Team

• Verry
• Richal
• Kiki

License

[Insert Appropriate License]

Disclaimer

Experimental emotion detection system. Not a substitute for professional psychological assessment.

Model

Link model nlp_emotion_indobert.h5 : https://drive.google.com/drive/folders/1-ngrmQ4lJyGrkOp2UWqhx2qeITodD4Za