Text-Based Emotion Detection and Applications: A Literature Review

Emotion Detection is the sentiment analysis process used to extract emotions from the text that best represent the author's mental state. In recent years, the emotion detection domain has become very popular due to its potential applications in artificial intelligence, human-computer interaction, psychology, and marketing. Despite its vast application, emotion detection is challenging in natural language processing. Detecting emotions from a text or image requires exhaustive knowledge and analysis. However, with machine learning, artificial intelligence, and data mining advancement, it has become possible to face this chal-lenge. Furthermore, the huge amount of textual data available online through social media, blogs, news, and articles helped the cause. For emotion detection, most of the studies have relied on machine learning and deep learning models and achieved good results. However, the researchers face some challenges that need to be addressed, such as the inability to extract the semantic information, the feature extraction process being time-consuming and inefficient, difficulty in identifying different emotions from non-standard language, imbalanced datasets, etc. This article aims to explore the existing approaches, methods, and evaluation measures used for emotion detection. The significant contribution, the methodology applied, and the results obtained by different researchers to gain the best possible results are also highlighted. Finally, the article highlights the limitations and provides the future direction that can be useful for research in emotion detection from text.