Chasher-Disha (চাষের দিশা): Bangla NLP Chatbot

A Bengali language chatbot built with Django and Bangla-BERT embeddings, capable of understanding natural Bangla input and responding intelligently. Users can interact through a web-based frontend.

Features

• Semantic understanding using Bangla-BERT (sagorsarker/bangla-bert-base)
• Handles multiple Bangla variations for the same intent
• Live chat interface with AJAX frontend
• Easy to extend by updating intents.csv
• Fallback response for unknown queries
• CPU-friendly for VPS or local development

Demo

• Type a message in Bangla in the chat box.
• Bot responds using semantic understanding.
• Example inputs:
  • "তুমি কে"
  • "হ্যালো"
  • "আজকের আবহাওয়া কেমন"

Installation

1. Clone the repository

git clone https://github.com/yenHunter/Chasher-Disha.git
cd Chasher-Disha

2. Create a virtual environment

python -m venv env
source env/bin/activate       # Linux/Mac
env\Scripts\activate          # Windows

3. Install dependencies

pip install -r requirements.txt

4. Create a .env file in root Get your Open Weater API key from OpenWeather

# .env 
OPENWEATHER_API_KEY=your_api_key_here

5. Run the server

python manage.py runserver

6. Open in browser

Visit http://127.0.0.1:8000 to chat with the bot.

Usage

• Type Bangla messages in the chat box.
• Bot will reply based on semantic similarity.
• Add more phrases or responses in intents.csv to expand capabilities.

Customization

• Each intent in intents.csv has:

  • intent: unique name
  • examples: semicolon-separated example phrases
  • responses: semicolon-separated responses

• The chatbot automatically computes embeddings for all examples.

License

MIT License

This version includes:

• Python & Django badges
• Placeholder for demo GIF or image
• Clear sections for Features, Installation, Usage, and Customization

About me

I’m a passionate web developer with experience in building scalable and secure applications. I specialize in Laravel, Django, JavaScript and server management with Nginx. I enjoy automating workflows, optimizing performance, and deploying modern web solutions on VPS environments. I believe in clean code, continuous learning, and sharing knowledge through open-source contributions.

Connect & Support