Abstract: The Web site is now a day full of information about people's opinions through apps such as social media, microblogging sites, review websites, personal blogs, etc. Sentimental analysis is an area where people's opinions may be evaluatedand categorized as positive, negative or neutral in text mining. In this paper, the tweets or review feelings published on theTwitter are recognized by searching the specific term in tweets and then the polarity of the tweets is assessed as positive andnegative. The tweet emotions are tweeted on a Twitter depending on the selection of the features of each score. Naive BayesClassifier (NBC) is used for training and testing word features, as well as assessing the feeling polarity of each tweet. tochoose the best features. Parameters such as accuracy, precision and time are taken into account in assessment ofperformance compared to three machine classifications, namely the Random Forest, Naive Bays and Support VectorMachines (SVM).Keywords: Support Vector Machine (SVM); Naive Bayes Classifier (NBC); Twitter Analytics;Sentiment analytics