Analyzing sentiment system to specify polarity by lexicon-based

Currently, sentiment analysis into positive or negative getting more attention from the researchers. With rapid development of internet and social media have made people express their views opinion publicly. Analyzing in impact many fields such as services productions that companies offer. Movie reviewer needs processing to be prepared detect emotion, classify them achieve high accuracy. The difficulties arise due structure grammar language manage dictionary. We present a system assigns scores indicating each distinct entity text corpus. Propose an innovative formula compute polarity score for word occurring find it dictionary we remove text. After classification, words are stored list will used calculate results reveal achieved best accuracy 76.585%.