johannesgirmaw/EthioMart_Amharic_NER

# EthioMart Amharic NER Project This project aims to build a Named Entity Recognition (NER) system for Amharic text from Telegram e-commerce channels to identify product names, prices, and locations. It leverages large language models (LLMs) and advanced NLP techniques. ## Project Structure The project follows a standard data science project structure. Refer to the folder structure for detailed organization. ## Setup and Running 1. **Clone the repository:** `git clone ` `cd EthioMart_Amharic_NER` 2. **Create a virtual environment (recommended):** `python -m venv venv` `source venv/bin/activate` (Linux/macOS) `venv\Scripts\activate` (Windows) 3. **Install dependencies:** `pip install -r requirements.txt` 4. **Configure environment variables:** Create a `.env` file in the root directory and add your Telegram API `api_id` and `api_hash`. ```env TELEGRAM_API_ID=YOUR_API_ID TELEGRAM_API_HASH=YOUR_API_HASH ``` 5. **Run the data ingestion script:** `python src/data_ingestion/telegram_scraper.py` Further instructions will be added as the project progresses.