This paper presents a comprehensive survey of emotion analysis (EA) in natural language processing (NLP), highlighting trends, gaps, and future directions. The authors analyze 154 relevant NLP publications from the last decade to address four key questions: (1) How are EA tasks defined in NLP? (2) What are the most prominent emotion frameworks and which emotions are modeled? (3) Is the subjectivity of emotions considered in terms of demographics and cultural factors? (4) What are the primary NLP applications for EA? The survey reveals that EA tasks are often defined using terms like emotion detection, classification, recognition, and prediction, with emotion recognition being the most common. The most widely used emotion frameworks are Ekman's basic emotions and Plutchik's model, which categorize emotions into six and eight basic types, respectively. However, these frameworks do not account for the subjectivity of emotions in terms of demographics and cultural factors, leading to a lack of diversity in emotion labels and annotation schemes. The survey also highlights the lack of standardized terminology in EA, which hinders gap identification, comparison, and future goals. Additionally, the absence of interdisciplinary research isolates EA from insights in other fields. The paper discusses the importance of considering demographic and cultural factors in EA, as well as the need for more diverse emotion categories and frameworks. It also emphasizes the importance of interdisciplinary research and the need for more inclusive and diverse systems in EA. The paper concludes with recommendations for future research, including the need for more diverse datasets, the development of more flexible and adaptable emotion analysis models, and the importance of interdisciplinary collaboration in EA.This paper presents a comprehensive survey of emotion analysis (EA) in natural language processing (NLP), highlighting trends, gaps, and future directions. The authors analyze 154 relevant NLP publications from the last decade to address four key questions: (1) How are EA tasks defined in NLP? (2) What are the most prominent emotion frameworks and which emotions are modeled? (3) Is the subjectivity of emotions considered in terms of demographics and cultural factors? (4) What are the primary NLP applications for EA? The survey reveals that EA tasks are often defined using terms like emotion detection, classification, recognition, and prediction, with emotion recognition being the most common. The most widely used emotion frameworks are Ekman's basic emotions and Plutchik's model, which categorize emotions into six and eight basic types, respectively. However, these frameworks do not account for the subjectivity of emotions in terms of demographics and cultural factors, leading to a lack of diversity in emotion labels and annotation schemes. The survey also highlights the lack of standardized terminology in EA, which hinders gap identification, comparison, and future goals. Additionally, the absence of interdisciplinary research isolates EA from insights in other fields. The paper discusses the importance of considering demographic and cultural factors in EA, as well as the need for more diverse emotion categories and frameworks. It also emphasizes the importance of interdisciplinary research and the need for more inclusive and diverse systems in EA. The paper concludes with recommendations for future research, including the need for more diverse datasets, the development of more flexible and adaptable emotion analysis models, and the importance of interdisciplinary collaboration in EA.