Study on preprocessing phase of Sarcasm Detection model in Social Media Text Conversation Using Support Vector Machine

Abstract

Sentiments plays very important role in human’s life. One of the most important sentiments is Sarcasm. This paper focus on the study of very first phase i.e preprocessing, on the detection of Sarcasm. Sarcasm is a type of verbal irony which includes telling something with a mockery or sarcastic tone that is just opposite to what actually is meant. It is basically a form of expression that is used show ridicule or amusement.

Sarcasm often relies on tone of voice, facial expressions, and context to be properly understood, which can make it challenging to detect in written text alone. In text conversations, sarcasm is sometimes indicated through the use of exaggerated punctuation (such as "Oh, great."), emojis, or explicit markers like "I'm just thrilled" to help convey the intended ironic tone. So, basically our model is designed to detect sarcastic comments via twitter data with the help of SVM. So this paper focuses on the preprocessing or feature extraction of textual data received from social media platforms.