Aspect based summarization of reviews using na?ve Bayesian classifier and fuzzy logic

E-commerce is getting popular, more and more products are sold online everyday and product reviews are growing rapidly. The larger number of reviews makes it difficult for customers and manufacturers. For popular products there may have thousands of reviews. Customers and manufacturers may not be able to understand overall opinion about aspects of products without going through all reviews. Also most of the existing methods make use of approximate summarization of product reviews. Here a more precise and realistic value of opinion is retrieved through naïve Bayesian classifier and fuzzy method. In this, in addition to the identification of opinions, we also identify linguistic hedges and apply fuzzy rules to magnify the effect of opinion.