Sarcasm and Humor Detection in Code-Mixed Hindi Data: A Survey

Sarcasm and humor play vital roles in natural language, facilitating human expression and social interaction. In code-mixed languages like Hindi, where multiple languages intertwine, detecting humor and sarcasm poses unique challenges for researchers and developers. This paper provides a comprehensive survey of sarcasm and humor detection, focusing on code-mixed Hindi and English data. It explores linguistic complexities, emphasizing diverse features contributing to humor and sarcasm. Machine learning and deep learning techniques are examined, highlighting their relevance in code-mixed data contexts. The development of a code-mixed Hindi humor and sarcasm dataset is discussed for training and evaluating detection models. The paper critically analyzes existing research, experimental results, and practical implications, addressing challenges and future directions. It aims to bridge the gap between evolving code-mixed languages and the need for accurate humor and sarcasm detection. Further exploration and advancement in this domain are encouraged.