With the increasing use of Twitter, social media that works in real-time for the public can convey complaints and appreciation to airlines, it is necessary to create a system that can classify a tweet containing opinions including what is the best class, in this study there are positive and negative classes. This is done so that it can help airline companies in terms of evaluating service improvements and can help people choose the right airline. Thus a sentiment classification with Lexicon Based features which is able to receive information in languages other than Indonesian (in this study used in English) is done to conduct sentiment analysis. Use the support vector machine algorithm to classify. The results of this study show optimal parameters and the effect of using Lexicon Based Features. By using parameter C is 10 and the learning rate is 0.03 also used Lexicon Based Features with an iteration of 50 times giving accuracy 40%, precision 40%, recall 100%, and f-measure 57,14%.