Gerding, Enrico, Larson, Kate and Jennings, Nick
Eliciting Expert Advice in Service-Oriented Computing
Agent-Mediated Electronic Commerce: Designing Trading Strategies and Mechanisms for Electronic Markets, Lecture Notes in Business Information Processing, Volumy 59.
We consider a setting where a consumer would like to elicit independent but costly reports from third-party experts about the reliability of a number of service providers. These reports can be of variable accuracy, but more accurate reports will be more costly to produce. The consumer can fuse reports from several experts to choose the provider with the highest probability of success. The goal in this paper is to find a mechanism which incentivises the experts to truthfully reveal the accuracy of the reports, and to induce the experts to invest costly resources in order to increase this accuracy. The challenge in doing so is that, while we can verify the success or failure of the selected service provider, we have no feedback about those service providers which were not selected. Moreover, we need to determine how to reward individual experts when the choice of service provider is based on a fused report from all exeperts. We explore a number of mechanisms to address this setting, including scoring rules, and indicate the problems in obtaining both truth telling and inducing the experts to produce accurate reports. We present a partial solution to this problem, and discuss remaining challenges.
Actions (login required)