Social media platforms such as Twitter have turned into key spots for folks to share their feelings and views over the last few years. It helps researchers and businesses get a better grip on what users think about stuff like events, products, or services when they dig into that kind of data. This study basically aims to compare different machine learning methods for sorting out sentiments in Twitter posts. They look at how well models like Logistic Regression, Naive Bayes, Support Vector Machine or SVM, Random Forest, and XG Boost do at labelling tweets as positive, negative, or neutral. For preprocessing the data, they use the Natural Language Toolkit, you know, things like lemmatization, tokenization, and cleaning up the text. Then to turn all that text into numbers, they apply feature extraction steps such as removing stop words and using the WordNet Lemmatizer.