METHODS OF EXTRACTION AND ANALYSIS OF PEOPLE'S SENTIMENTS FROM SOCIAL MEDIA

Abstract

This study explores methods for extracting and analyzing people's sentiments from social media, utilizing advanced natural language processing and machine learning techniques. The goal is to recognize sentiments and based on them to categorize posts and comments as positive, negative, or neutral in order to understand users' attitudes and emotions. The algorithms used for sentiment classification in this research include Naive Bayes, SVM, Logistic Regression, and Random Forest. Additionally, the study analyzes and compares the performance of these algorithms in terms of accuracy, recall, and F1 score, providing a comprehensive overview of their effectiveness. It also emphasizes the importance of hyper-parameter tunning to improve the accuracy of classification algorithms.
The results of this study can be used to assist social media platforms, researchers, and policymakers in developing strategies to manage and improve user experiences, as well as to make informed decisions based on user reactions.