Cooperativeness prediction in P2P networks
Expert Systems with Applications: An International Journal
A robust methodology for prediction of trust and reputation values
Proceedings of the 2008 ACM workshop on Secure web services
Managing data for evaluating trust in unstructured peer-to-peer networks
GPC'08 Proceedings of the 3rd international conference on Advances in grid and pervasive computing
Finding reliable recommendations for trust model
WISE'06 Proceedings of the 7th international conference on Web Information Systems
| Hi-index | 0.00 |
Current centralized trust models are inappropriate to apply in a large distributed multi-agent system, due to various evaluation models and partial observations in local level reputation management. This paper proposes a distributed reputation management structure, and develops a global reputation model. The global reputation model is a novel application of neural network techniques in distributed reputation evaluations. The experimental results showed that the model has robust performance under various estimation accuracy requirements. More important, the model is adaptive to changes in distributed system structures and in local reputation evaluations.