Trust Model Based On Islamic Business Ethics and Social Network Analysis

Abstract

Buyers and sellers in e-commerce market such as e-auction form a virtual community. They use the feedback system to rate each other following a completed transaction and these ratings are used to build their reputation in the virtual community.  Existing reputation systems can often be easily manipulated by forming cohesive group in giving fake user feedbacks to increase their respective reputation. This practice is a clear violation of Islamic business ethics. In addition, there is currently no real-time support for reputation system and this causes users to be misinformed on the reputation of a seller. To improve the reputation system this study developed a trust framework based on business Islamic ethics. In this paper, a trust model which evaluates conformance to nine Islamic business ethical codes is proposed to calculate users’ initial trust value based on their ethical behavior. The trust model proposed the Islamic business ethics algorithm which calculates the user compliance to Islamic business ethics (IBE) score based on trading partner’s feedbacks. Because of feedback frauds can still occur, this study introduces a cohesive group algorithm to track users who collaborate to give false feedbacks. The cohesive group algorithm applied k-core algorithms which is capable of determining the strength of the relationship of every user in the cohesive group. The cohesive group algorithm also proposed a cohesive score to determine the feedback reliability of every user’s transaction based on the user’s k-core and the highest k-core. In the group user reputation (trust score) is measured by considering the feedback reliability status for all transactions. A reputation prototype system for e-auction was developed as the test bed to validate the trust model through simulation of the set of initial experiments, showing the feasibility and benefit of the model.