A survey of sentiment analysis and opinion mining using supervised machine learning

Abstract

This paper presents a comprehensive research of sentiment analysis and opinion mining, focusing on the application of supervised machine learning techniques. The computational study of people’s thoughts, feelings, and views as they are expressed in written language is called sentiment analysis, sometimes referred to as opinion mining. In a variety of contexts, including social media, product reviews, and news articles, supervised machine learning—which necessitates labeled training data—has demonstrated great potential in precisely categorizing sentiment and opinion. Support Vector Machines (SVM), Naive Bayes, and deep learning models are among the classification and regression algorithms examined in the survey, which emphasizes how well they perform in sentiment categorization tasks. Analyzing various classifiers, such as Guileless Bayes, Multinomial Credulous Bayes, Bernoulli Gullible Bayes, Stochastic Inclination Plummet (SGD) Classifier, Nu SVM/Nu SVC, Direct SVM/Straight SVC, and Calculated Relapse, is at the heart of the investigation. Calculated Relapse arises as a hearty and powerful classifier across dataset sizes, while Direct SVM/Straight SVC shows a reliable expansion in exactness rates as dataset size develops. The discoveries help in informed decision-production for classifier determination and organization, improving the viability of sentiment analysis strategies across different areas.