Analysing Politician Tweets: Hybrid K-Naïve Bayes Sentiment Analysis Approach

Sentiment analysis is about identifying positive and negative sentiments in textual data using natural language processing (NLP). Companies frequently utilise it to monitor social media sentiment, evaluate brand reputation, and comprehend their clients. Sentiment analysis primarily focuses on a text’s polarity—positive, negative, or neutral—but it also goes beyond polarity to pinpoint specific moods and emotions, such as anger, delight, or sadness, as well as urgency—urgent compared to not urgent—and even intentions—interested instead of uninterested. Automated processes must be used to extract sentiments from objects on the Internet. In this paper, a hybrid model of K-Naïve Bayes classifier has been proposed which will categorise politician’s tweets seen on Twitter social media platform into six different categories, i.e. beginner positive, positive, neutral, beginner negative, negative, and compound. The accuracy, precision, and recall achieved by the proposed model are 89.9%, 85.5%, and 87.5%.