esmael-uta/EthioMart_NER

EthioMart_NER is an AI-powered solution designed to revolutionize Telegram-based e-commerce in Ethiopia. # EthioMart_NER # Business Need EthioMart aims to become the primary hub for all Telegram-based e-commerce in Ethiopia. With many independent Telegram e-commerce channels operating separately, customers and vendors face challenges managing multiple channels for product searches, order placements, and communication. EthioMart plans to address this by creating a centralized platform that aggregates real-time data from various e-commerce Telegram channels into one unified channel, offering a seamless experience for both vendors and customers. ## Project Focus Fine-tuning LLMs for Amharic Named Entity Recognition (NER). The NER system will extract key business entities (like product names, prices, and locations) from: - Text - Images - Documents Shared across various Telegram channels. This extracted data will feed into EthioMart's centralized database, transforming it into a comprehensive e-commerce hub. ## Key Objectives ### Real-time Data Extraction from Telegram Channels - Automatically gather and update product listings, prices, and other relevant information from different Telegram e-commerce channels. ### Fine-tuning LLMs for Amharic NER - Customize and optimize large language models to recognize and extract relevant entities from the content shared in Amharic. ## Entity Extraction Targets - **Product Names or Types**: Identify the names and types of products being sold. - **Materials or Ingredients**: Extract mentions of specific materials or ingredients in product descriptions. - **Location Mentions**: Detect geographic locations related to product availability or delivery. - **Monetary Values or Prices**: Extract pricing information and related monetary values.