Emotion Detection from Text and Analysis of Future Work: A Survey

Abstract

Emotion detection has recently become a fascinating subject for academics, as people express their emotions in various ways, including through manners, words, writing, or facial expressions. With the increased use of social media and technological advancements, it has become easier to retrieve enormous emotions regularly. Unsurprisingly, social networks have become a suitable communication medium for users to interact with friends and share their emotions. Detecting emotion is a valuable and powerful tool for identifying and recognizing moods, with numerous uses that can impact diverse areas. Machine learning algorithms and approaches are used to find high-quality solutions for detecting emotional issues among social media users and different datasets. This survey focused on emotion detection and classified the most current state-of-the-art systems for textual emotion recognition based on methodology, emotional model, and several datasets. Our Conclusion aims to emphasize the limitations and gaps in these recent efforts, analyze the approach and detection technique, and suggest future research directions to fill these gaps in this rapidly evolving field. We highlighted different approaches to detecting emotion, including the current trending successful approach, and conducted a comparative study to reveal the algorithm's performance using other models, approaches, and methodologies.