Sarcasm Detection Using Deep Learning
Deepak Sahoo, Nayan Ranjan Paul, Rakesh Chandra Balabantaray, Asif Uddin Khan · 2021
In social media, people use positive words or give more stress on positive words to express their negative opinion which is known as sarcasm. This on purpose ambiguity makes sarcasm detection is a challenging and popular task in natural language processing. In this work, we have used a CNN-GRU based model with word embedding to find the occurring of positive words. In evaluation metric we have used precision and recall to measure the accuracy of the CNN-GRU based model. We achieved 72.58% precision and 83.06% recall. The proposed system is compared with some existing system and the system gives encouraging results.