Compare and Evaluate Pre-Trained Models for Sentiment Analysis

[abccastro]

Finalize the selection of the pre-trained model for analyzing sentiment in flight reviews by comparing and evaluating the accuracy of results from both RoBERTa and Vader models.

[abccastro]

As discussed with the team, Vader model will be used for our chatbot sentiment analysis as it gives a higher accuracy compared to RoBERTa. See evaluations below:

• https://github.com/abccastro/ChatBot-Online-Flight-Booking/blob/main/Flight%20Sentiment%20Analysis%20Dataset%202.ipynb
• https://github.com/abccastro/ChatBot-Online-Flight-Booking/blob/main/Flight%20Sentiment%20Analysis.ipynb