رضوان محمدی باغملایی

This survey explores the role of emojis in improving sentiment analysis, emphasizing their importance in capturing subtle emotions in digital communication. It reviews methods from lexicon-based and machine learning approaches to advanced deep learning and transfer learning models, showing improved accuracy in emotion detection across diverse datasets like Twitter. The study discusses challenges such as emoji polysemy, cultural differences, and dataset imbalances, while also examining real-world applications in healthcare, marketing, and education. The survey also highlights potential directions for future research, including the development of culturally diverse datasets and scalable multimodal systems for more context-aware sentiment analysis.