Reputation based trust models are widely used in e-commerce applications. The feedback comments are used to compute sellers’ reputation trust score. This system is based on the observation that the buyers can express their opinions openly and honestly in free text feedback comments. If there are no feedback comments available then the buyer need to consider the features specified for each product. The tool SentiWordNet is used for the extraction of feedback comments into positive, negative and neutral. K-means clustering method is used to group the data obtained after sentimental analysis. Each sentence in a feedback comment is considered as a document. This calculation is lead to obtain a sellers trust profile. The problem faced by all the reputation system is an all good reputation problem where reputation score are universally high for sellers and this will be difficult for a potential buyer to identify the potential buyer. This system will provide a solution for this system.