Relation between the amount of Sarcasm and the Ratings of English TV shows

Sarcasm detection is one of the most challenging parts of Natural Language Processing (NLP) since it involves understanding the context in which a particular text is delivered. In recent times there has been much advancement in Machine Learning (ML) technologies that can be leveraged to detect sarcasm in a text. Popular English TV shows of the genre of Situational Comedy involve sarcasm or satire in their dialogues. This study proposes that the ratings of the Popular English TV shows are dependent on the sarcasm content in that show and thus incorporates a classifier model first to detect whether sarcasm is present in the dialogue and then proceeds with a statistical test to determine whether there exists a relationship between sarcasm and the ratings provided to the show. This work can provide value to various content platforms for their recommendation systems as well as give insights to content creators for staying competitive in the business. The results of the statistical test are interpreted, and a scope for further investigation is discussed.