This study focuses on the analysis of tweets, which are textual exchanges among users on the social media platform Twitter, serving as a platform for expressing opinions on various subjects. With 321 million Twitter users and approximately 6000 tweets generated per second, sentiment analysis becomes crucial in understanding the emotional context of these messages. Sentiment analysis, a facet of natural language processing, is employed to determine the emotional tone, categorizing content as positive or negative. The unstructured text for sentiment analysis is sourced from diverse online platforms like blogs, reviews, and comments. Through automatic, hybrid, or rule-based cleaning processes, the text is prepared for the subsequent stage. The evaluation of emotional content is conducted using algorithms derived from artificial intelligence, data mining, and machine learning. This research integrates natural language processing and machine learning techniques for sentiment analysis, employing methods such as Document Term Matrix. The study utilizes supervised machine learning algorithms, including Random Forest, Poisson, Multivariable, Gaussian, Bernoulli, and Simple Naive Bayes, for text classification