Comparative Analysis of Fine-Tuned LLM, BERT and DL Models for Customer Sentiment Analysis

The fine-tuned Large Language Models such as Generative Pre-Trained Transformers (GPT), Google's and BERT models are leveraged for NLP tasks. The online businesses are relying on customers' online review posting and positive feedback to improve the products sales and services. To predict the accurate emotion and sentiment of the customers remains challenging. There are literatures and sentiment models' research studies show that the traditional models are having performance and accuracy issues. To overcome the sentiment prediction accuracy and model performance issues, author propose the fine-tuned Mistral LLM and BERT models. These models' experimental study results show that fine-tuned LLM outperforms traditional models, and it predicts more accurate sentiments of the customers.