Emotion Detection with Transformers: A Comparative Study

Analyzing the sentiment and emotion behind social media text data can help us understand the attitudes, preferences, and feelings of the users (Isah, Trundle, & Neagu, 2014). Unveiling these emotions goes beyond simply gauging sentiment, as it provides deeper insights into user motivations and psychological states. For instance, capturing the nuances of human emotion expressed through text remains a complex task due to the limitations of language itself. Sentiment analysis, while valuable, only provides a surface-level understanding. Identifying specific emotions expressed in text offers deeper insights into user behavior and motivations. Sentiment analysis is a natural language processing task that aims to classify the polarity of a text as positive, negative, or neutral (Khurana, Koli, Khatter, & Singh, 2023; Min et al., 2023). Emotion classification is a related task that aims to identify the specific emotion expressed in a text, such as sadness, joy, love, anger, fear, or surprise. Both tasks might be challenging due to the complexity and variability of natural language. Transfer learning is a technique that allows a model trained on one task to be used for another related task (Weiss, Khoshgoftaar, & Wang, 2016).