Systematic literature review on sentiment analysis using transformers

Key Points

• Sentiment analysis using transformers has grown significantly in recent years, offering improved accuracy and efficiency across various datasets.
• A key finding shows that transformer models outperform traditional methods in text classification tasks, particularly in understanding context.
• The analysis covers various transformer architectures and their applications in different sentiment analysis scenarios according to recent literature.
• This work emphasizes the need for further exploration of transformer adaptations to enhance sentiment analysis frameworks and methodologies.