sylvainsbrr / p7-api

Sentiment Analysis for Social Media - Air Paradis Project

Overview

This project aims to develop an AI prototype for Air Paradis to predict sentiments from tweets. Given the airline's interest in monitoring and improving its social media image, we focus on creating a robust, scalable solution that leverages advanced AI models to forecast potential PR issues effectively.

Objectives

• To create a prototype AI model that accurately predicts the sentiment of tweets.
• To prepare a methodology presentation for a non-technical audience, detailing the "advanced custom model" approach.
• To explore and implement various modeling approaches including a simple custom model, an advanced custom model, and a BERT model.
• To integrate MLOps practices throughout the model development lifecycle for enhanced efficiency and reproducibility.

Project Structure

Air-Paradis-Sentiment-Analysis/
│
├── model/                 # Trained model files and versioning
├── scripts/              # scripts for mlflow prototyping
├── api/                   # api files
└── README.md               # Project overview and setup instructions