A Survey on Humor Detection Methods in Communications

Humor is a part of most of the communications these days. Modeling the humor present in verbal communication is trending recently. Assessing the degree of the humor present in text is a challenging task. Most of the existing studies on humor detection depended on binary classification with respect to the linguistic features. Deep Learning is the best alternative and most prominent technology that is most widely being used in humor detection in verbal communications. Humor Detection needs effluent models and patterns which are not available for all the cases. The researchers in this area are mushrooming day by day. As an effort to boost research and ignite new ideas in this challenging area, this paper explores how deep learning is being used in humor detection. Also reviews various humor detection methodologies which are based on deep learning. This paper presents the critical analysis of the humor detection methodologies and future directions.