Support Vector Machine Classifier with Principal Component Analysis and K Mean for Sarcasm Detection

Sentiment analysis is a task which evaluates an opinion expressed is positive, negative or neutral but presence of sarcasm flips the polarity of the text. Various different machine learning algorithms can be applied for the detection. The various phases are applied for the sarcasm detection like data set collection, feature extraction and classification. A feature reduction method can also help in improving the classification accuracy of the classifier. We have compared the performance of various classifiers- Naïve Bayes, support vector machine, K-nearest neighbour, support vector machine with principal component analysis and support vector machine along with principal component analysis and K-means clustering for sarcasm detection on the twitter dataset. As compared to other classifiers, the combination of K-mean, PCA and SVM provides high performance in terms of accuracy.