Logo Lanfrica

BekiChemeda/Amharic-pos-tagger-telegram-bot

Domaine:

natural language processing

Type de record:

software
Créateur:
Bek
Hôte:
# Hayyuu Amharic POS Tagger Bot Hayyuu is a Telegram bot designed to perform Part-of-Speech (POS) tagging for Amharic and other Ethiopian languages using a Conditional Random Fields (CRF) model. ## Features - **POS Tagging**: Analyzes Amharic sentences and tags each word with its part of speech. - **Custom Tokenizer**: Specifically designed for Amharic, handling special characters and punctuation. - **Telegram Bot Commands**: - `/start`: Start the bot and get a welcome message. - `/about`: Learn about the bot and its development. - `/help`: Get a guide on how to use the bot. - `/info`: Understand more about NLP and POS tagging. - `/teach`: (Coming soon) Help the bot learn new data. ## Installation 1. **Clone the repository**: ```bash git clone github.com cd Amharic-pos-tagger-telegram-bot ``` 2. **Install dependencies**: ```bash pip install -r requirements.txt ``` 3. **Set up your channel**: - Update the channel username in the code: - Open `hayyuu.py` and replace `@Bright_Codes` with your own Telegram channel username. - Make the bot an **administrator** in your channel: - Go to your Telegram channel settings and add your bot as an admin. Ensure it has the necessary permissions to check user membership. 4. **Set your API token**: - create .env file and add your API token you got from bot father bot to it. ```bash API_TOKEN=Your_Bot_API_YOU_GOT_FROM_BotFather ``` 5. **Run the bot**: ```bash python hayyuu.py ``` ## How to Use 1. **Start the bot**: - Open Telegram and start a chat with your bot. - Send `/start` to initialize. 2. **Interact with the bot**: - Use commands like `/help`, `/about`, and `/info`. - Send an Amharic sentence to see POS tagging in action. ## Requirements Ensure you have Python 3.8 or higher installed. Install the dependencies from `requirements.txt` to set up the environment. ## Model Information The bot uses a CRF model trained on Amharic data. The model file `tuned_cr …