In the era of Web 2.0 e-commerce website plays a very important role in our everyday life. If the consumer wants to purchase a product from the e-commerce website then reviews plays a very important role in decision making about a product. The reviews are unstructured reviews because there is no fixed format for customer to write the reviews. Because of these huge quantities of unstructured reviews there is a need to develop a technique which extract the features from the reviews and gives the opinion about the features. In this proposed technique i.e Features Based Opinion Classification (FBOC), identification of features is done, then after the identification, feature is classified as according to their opinion whether positive, negative and neutral using two methods namely AFINN and VADER (Valence Aware Dictionary for sEntiment Reasoning). Summary of feature is generated showing total number of positive, negative and neutral opinion of features and visualized in a form such that it can be easily understandable by end user.