Crowdsourcing has developed as a significant problem solving and information gathering standard which makes use of the Internet. The crowdsourced data available on different blogs, forums and social media websites in the form of customer feedback and reviews is an important source of information for the companies. A Data analyst aims at finding trends, patterns and meaningful insights from crowdsourced data. Taking help of crowdsourced workers to analyze this redundant and huge amount of data is not only tedious but is also time consuming. A solution for this problem is proposed that increases the potential of gain of meaningful insights to a large extent. Data mining approach is used to analyze reviews of four different ecommerce websites based on different performance parameters. This unstructured crowdsourced data is processed and filtered using J48 classifier. 355 fake reviews were eliminated. Remaining truthful data was analyzed and Multiviewpoint based similarity clustering was used in order to remove redundancy, which gave the accuracy of 98.66%. To make crowdsourced data more representative and usable graphs were plotted based on delivery, return, refund and customer care parameter to help the analyst in decision making