ANALYZING SARCASM IN TWITTER DATA USING MACHINE LEARNING ALGORITHUM

ANALYZING SARCASM IN TWITTER DATA USING MACHINE LEARNING ALGORITHUM

MS. NAMRATA GOKUL WAGH1, PROF. HARISH BARAPATRE2

1 Research Scholar, SES, Group of Institute, Bhivpuri Road Karjat (MS), w.namrata3@gmail.com, University of Mumbai.

2 Professor, Yadavrao Tasgaonkar Institute, Bhivpuri Road Karjat (MS), harishkbarapatre@gmail.com, University of Mumbai.

Abstract: - It seems like We're providing information about sarcasm, sentiment analysis, and the evaluation of various machine learning algorithms for sentiment analysis. We text touches on how sarcasm is a form of expression where people say the opposite of what they mean, often to convey criticism. We also mention that sarcasm can be difficult to detect due to its obscurity and that sentiment analysis can be improved by analyzing and interpreting sarcasm in sentences. We go on to discuss Twitter as a platform where people express their thoughts and emotions in real-time, highlighting its significant volume of information and the potential challenges related to accuracy and trustworthiness since anyone can post anything. We mention the idea of Twitter battles where individuals attack each other. Lastly, we mention the evaluation of several machine learning algorithms (Random Forest, Naive Bayes, Bayesian Network, Support Vector Machine, K-Nearest Neighbour, Multi-Layer Perceptron) and a deep learning algorithm (Recurrent Neural Network) for sentiment analysis. According to We are evaluation, Random Forest performs the best in terms of prediction accuracy and error rate.

Keywords: K-nearest neighbours (KNN), Random Forest (RF), Naive Bayes (NB), Application programming interface (API), Online social networks (OSNs).