Sarcasm Detection in News Headlines using Voted Classification
Santosh Kumar Bharti, Rajeev Kumar Gupta, Nikhlesh Pathik, Ashish Mishra · Proceedings of the 2022 Fourteenth International Conference on Contemporary Computing · 2022
Sarcasm is a sentiment one uses to advocate the opposite of what they mean. Sarcasm is purely context-based, a common phenomena in social media and is inherently difficult to detect, which makes it sometimes difficult for humans to interpret. Over the past couple years, studies on sarcasm detection have been mainly focused on social media content or review analysis which are usually noisy in terms of labels and language. To overcome those issues hence, this paper deals particularly with sarcasm detection in News Headlines. The approach implemented is bag of words analysis using term frequency and n-grams frequency followed by voted classification. We consider comparing different features-based approach and the experimental results are generated by a voted classifier consisting of 7 different classifiers. Result metrics are included in the paper such as accuracy, precision and recall to give a better picture of how efficient the model is.