Product reviews have now become a key to business intelligence. All the products available online now provide a customer review and rating of their products for quick decisions and comparisons. Apps, software, products and even websites are reviewed by the customer for their rating among a variety of other products. Different blogs and websites like ?WhatMobile?, and ?gsmarena? use formal ways to review like survey forms and star rating. Social media is being used for sharing personal experiences with different products and these public reviews also play an important part in the business performance. Sentiment Analysis is a way of recognizing and classifying opinions spoken in a piece of text, particularly in order to check the user''s behavior about some matter, item or issue. This study is based on the extraction of reviews tweeted by the twitter users about their mobiles. This study highlighted the way of expressing their views in a good way by analyzing their comment and mining the opinion. E-trade and social media have changed the level of online customers and it is a regular practice for them to look at the rating of items before making a buy. For this study, Twitter Social media data set is used for getting user opinion from this free social media site as Twitter is a social media microblogging platform of users for sharing views through text containing emoticons and hashtags. This study recommends fuzzy classification of opinion mining system that is mostly binary but this work showed multilevel sentiment analysis of user tweets. Sentiment analysis was a stepwise process that included preprocessing, Part of Speech (POS) tagging, separating these parts as per our requirement, and finally showing the polarity of that sentence by using fuzzy logic approach. The proposed strategy broadens the feature-based arrangement approach and also specifies the strength of emotions related to the tweets by using modifiers, emoji, and concentrators. This research work used the Fuzzy classifier in a deeper sense by depicting the opinion in these 5 levels i.e. excellent, good, average, bad, and very bad. This 5 level fuzzy classification enhanced the vision of opinion more clear and accurate than bipolar opinion.