Publication

Abstract : At present, Micro-blogging has enhanced an integral part of modern life, allowing individuals and organizations to communicate their thoughts and opinions in a variety of ways. Due to the rise of social media platforms, people from various countries have started using blogs and Twitter to share their thoughts and feelings which are expressed as tweets. The task of recognizing the perspective or an attitude of the user that is represented on a topic and is called sentiment analysis. Sentiment analysis is used to designate text or words into a diversity of sentiments. The paper proposed the method for classifying the polarities of sarcastic sentences. Initially, data is pre-processed and feature extraction is done that includes tokenization, lemmatization. To withdraw the word frequency is the most primary form of analyzing textual data. The one only sentence is tiny for an entity for noticing the word distribution. Therefore, the evaluation of the word frequency is done on all positive sentences. Further, we use machine learning for identifying polarities of sarcasm in the sentence. Polarities are deliberated for the patterns based on their structure and the features of phrases in the sentences. The experimental results are favorable with good accuracy for polarities of sarcastic sentences based on the performance metrics.