Exploring airline reviews using NLP models

This project will extract insights from user reviews of European airlines. The investigation comprises of two major exercises:

1. Get the user review data from Skytrax on the four major European airlines (code is in this notebook).
2. Import, clean, and perform the sentiment analyses using Natural Language Processing (NLP) modules in Python. The code is present in this notebook. We will cover four different NLP approaches and evaluate their performance:
   • TextBlob
   • VADER (Valence Aware Dictionary and sEntiment Reasoner)
   • RoBERTa - a pre-trained model, including the Huggingface pipeline
   • Flair framework - Read more information here.