Possibilistic logic for intelligent business agents


R. Lau, O. Wong




With the rapid growth of electronic commerce on the Internet, it becomes increasingly important to have effective, secure, and low-cost methods of handling the monetary aspect of the on-line transactions. Recently, a number of electronic payment services have been introduced to the Web. An agent-based approach for electronic payment is appealing since agents can proactively monitor the latest information about settlement services on the Internet, and autonomously select the best settlement option on behalf of their customers. However, because of the intrinsic dynamics of the Internet, these agents are faced with the challenge of making good decisions under the constraint of uncertain and incomplete market information. Possibilistic logic provides an expressive formalism to capture these uncertainties, and a robust and powerful reasoning method to make sound decisions by considering the uncertainties related to the activities in payment processing. In addition, possibilistic deduction can be used to explain and justify an agent’s decisions. Enhanced explanatory power promotes users’ trust and satisfaction, and this is essential in agent-mediated electronic commerce. This paper proposes an agent-based electronic payment service. In particular, how possibilistic: logic can be applied to develop intelligent business agents is discussed.