Abstract— Sentiment analysis, a subfield of natural language processing, plays a vital role in understanding public opinion and sentiment towards products, services, or events. In this study, we explore the field of sentiment analysis with a special emphasis on the use of machine learning techniques to classify the sentiments set in textual data. A Multinomial Naive Bayes classifier trained on a dataset of text data with sentiment markers is used in the study. The text data is pre-processed and vectorized using the Term Frequency-Inverse Document Frequency (TF-IDF) technique. This helps with a variety of operations, including request exploration, client feedback analysis, and social media monitoring. The study looks at the techniques used, assesses the effectiveness of the model, and addresses about the results and possible directions for farther sentiment analysis exploration.