A Comparison of Multinomial Naive Bayes and XG Boost for Sentiment Analysis and Bias Detection in Tweets

In the context of social platforms, this study explores Twitter sentiment analysis, providing practical insights and directions for future research. In previous days, people conveyed their feelings directly, but now platforms like Twitter have gained popularity for expressing individual and organizational opinions. Understanding these expressions can be challenging. This research focuses on sentiment analysis techniques for Twitter data, a subfield of Natural Language Processing (NLP) where machine learning plays a key role in detecting public emotions. The paper covers methods, tools, pre-processing, sentiment classification techniques, evaluation matrixes, and machine learning algorithms like Naive Bayes. The analysis, employing various keywords and hashtags, evaluates using metrics such as confusion matrix, F1 score, precision, and recall. The study emphasizes the responsible and ethical use of sentiment analysis tools for informed decision-making, promoting inclusivity and empathy online. Through systematic analysis, a million tweets were categorized into positive and negative sentiments, contributing to the growing body of knowledge in sentiment analysis.