Optimal Feature Extraction based Machine Learning Approach for Sarcasm Type Detection in News Headlines

Sarcasm detection has received increasing research in recent years. Detection of sarcasm is of great importance and beneficial to many NLP applications, such as sentiment analysis, opinion mining and advertising. Detection of sarcasm is of great importance and beneficial to many NLP applications, such as sentiment analysis, opinion mining and advertising. Generally, sarcasm detection task is treated as standard test classification problem. Sarcasm is the unconventional way of conveying a message which conflicts the context. It can lead to a state of ambiguity. As Sarcasm represents contrary sentiment to the literal meaning that is conveyed in the text, it is hard to identify sarcasm even for a human. Existing models mainly focus on designing effective features for improving the detection performance. In this paper, optimal features are to be selected before data passes to classification task. So data pre-processing makes the data clean so that the performance of the classifier will be enhance. Result shows the improve performance in sarcasm detection using the optimal feature sets.