Overview

I built a web interface for the real-time analysis of clinical texts. This tool helps visualize entity distribution, confidence scores, and entity maps, making it easier to process French medical records.

Key Features

• Real-time Analysis: Instant processing of clinical text inputs.
• Interactive Visualizations: Created dynamic charts for entity distribution and confidence scores using Matplotlib.
• Model Evaluation: Explored and evaluated multiple approaches, comparing traditional ML with Transformer models in a dedicated notebook.

Technologies Used

• Core: Python, Pandas
• AI/ML: DrBERT (Fine-tuned), Transformers
• Frontend: Streamlit