Logo Lanfrica

Kalze1/EthiopianMedicalDataWarehouse

Domaine:

healthcare

Type de record:

softwareproject
Créateur:
Kal
Hôte:
Building a data warehouse to centralize Ethiopian medical business data from Telegram and online sources. Develops ETL pipeline, data cleaning, object detection with YOLO, and a FastAPI interface for data exposure. Provides valuable insights into Ethiopia's medical business landscape. ## **Ethiopian Medical Data Warehouse** This project is a comprehensive pipeline for extracting, transforming, and exposing Ethiopian medical business data for analytics and object detection. The system leverages advanced tools like **Telethon**, **YOLOv5**, **DBT**, and **FastAPI** to automate data collection, transformation, and analysis. --- ### **Features** 1. **Telegram Data Scraping**: Collect data and images from Telegram channels using **Telethon**. 2. **Object Detection**: Use **YOLOv5** to perform object detection on scraped images. 3. **Data Transformation and Warehouse**: Transform and store cleaned data using **DBT** and PostgreSQL. 4. **API Exposure**: Build an API to access and manage data using **FastAPI**. --- ### **How the Project Works** #### **1. Data Scraping** - Telegram channels like `t.me` are scraped using the Telethon API. - Images and text data are saved locally for further processing. - **Tools Used**: Telethon, Python. #### **2. Object Detection** - YOLOv5 pre-trained models detect objects in scraped images. - Bounding box coordinates, class labels, and confidence scores are extracted. - Detection results are saved as a CSV and stored in a PostgreSQL database. - **Tools Used**: YOLOv5, PyTorch, PostgreSQL. #### **3. Data Transformation** - Raw data is transformed into analytics-ready datasets using **DBT**. - DBT models standardize and clean data for efficient querying. - Data is tested for quality and documented for clarity. - **Tools Used**: DBT, PostgreSQL. #### **4. API Exposure** - FastAPI provides RESTful endpoints to access and manage the medical business data. - Users can create, retrieve, and query data through the API. - **Tools Used**: FastAPI, SQLAlchemy, Pydantic. --- ### **Future Improvements** - Add user authentication to the API for secure data access. - Include more advanced data visualizations and analytics. - Automate data scraping and model updates with a scheduler like **Airflow** …

Languages