A New Feedback-Analysis based Reputation Algorithm for E-Commerce Communities.

Dealing with the ever-growing content generated by users in the e-commerce applications, Trust Reputation Systems (TRS) are widely used online to provide the trust reputation of each product using the customers' ratings. However, there is also a good number of online customer reviews and feedback that must be used by the TRS. As a result, we propose in this work a new architecture for TRS in e-commerce application which includes feedback' mining in order to calculate reputation scores. This architecture is based on an intelligent layer that proposes to each user (i.e. «feedback provider») who has already given his recommendation, a collection of prefabricated feedback to like or dislike. Then the proposed reputation algorithm calculates the trust degree of the user, the feedback's trustworthiness and generates the global reputation score of the product according to his «likes» and «dislikes». In this work, we present also a state of the art of text mining tools and algorithms that can