Sentiment Analysis Using Transformer Models (BERT, ALBERT, RoBERTa, and DeBERTa) on a SMILE Twitter Dataset

Sentiment analysis is essential for comprehending how people feel about different issues on social media sites. In this work, the effectiveness of utilizing transformer models, specifically BERT, ALBERT, RoBERTa, and DeBERTa, for sentiment analysis on a Twitter dataset. These models perform remarkably well in tasks involving Natural Language Processing because they have been pre-trained on enormous volumes of text data. By fine-tuning these models on a SMILE Twitter dataset, we aim to analyze and classify sentiments expressed in tweets containing various emotions. Through our experiments and evaluations, we demonstrate the capabilities and limitations of each model in accurately capturing sentiment nuances related to various emotions conveyed through tweets on social media. Our findings provide insights into the efficacy of these advanced language models for sentiment analysis tasks, paving the way for improved sentiment analysis techniques in social media monitoring and opinion mining.