Decision Theoretic Assessment Model for Online Business Games

Sanat Kumar Bista, Keshav Dahal, Peter Cowling


Several approaches to reputation and trustworthiness assessment based on probabilistic assessment have considered studying its usefulness in online business environments. It is seen that probabilistic models of reputation and trustworthiness assessment aids in resolving uncertainty by assessing the trustworthiness level of players. In this paper we intend to further enhance the assessment by combining the probabilistic assessment part with the expected utility of each player, thus resulting in a decision theoretic assessment. In this form of assessment, a player makes a decision on the basis of what it believes (given by the probabilistic assessment) and what it wants (given by the utility value of the choice). This method ensures that the players make a continuous measure of the state quality. The assessment of trustworthiness in this model is guided by the principle of maximum expected utility, which enforces a rational player to choose an action only if that meets its expected utility. Our results show that the decision theoretic models of assessment positively contribute the evolution of cooperation in a player society. Experiments have been carried out in a business game environment based on the principles of Iterated Prisoner’s Dilemma.  

Full Text: