A LEARNING-ENABLED INTEGRATIVE TRUST MODEL FOR E-MARKETS

Abstract

Existing e-markets presume no deception from agents or else they employ simple mechanisms to counteract deception. However, the reality shows that agents in e-markets can either cheat or break contracts due to higher benefits elsewhere, which is similar to what we find in humanity in general. Accordingly, the notion of trust in human society should be implemented in e-markets. Most of the existing research on trust is modeled theoretically from different views, and hence it is not easy to deploy them in e-markets due to the naturally non-computable essence of trust. However, current computable trust mechanisms, such as those used in eBay and Nextag, uniformly manipulate trust involved in all trading, resulting in complaints about non-differentiated experience. On the other hand, a computable trust model can help the formation of coalitions in e-markets and increase market competition. In this paper, we present a simple heuristic trust model absorbing the predominant views of trust with which agents in e-markets can better evaluate possible trading partners before trading processes take place. In this model, trust is characterized by the properties of being computable, individualized, evolutional, represented by scores, and extendable to the computation of coalition trust.