Skip to main content

Toward the next generation of computational trust and reputation models

Final Activity Report Summary - SOCIALREP (Toward the next generation of computational trust and reputation models)

The research in computational trust and reputation models is an area with a great relevance in the field of autonomous agents and multi-agent systems. These mechanisms, in a similar way of what happens in human societies, improve the security and robustness of the virtual societies. The last years, a lot of different models have been developed. However, till now there was no mechanism to compare these models in a common environment under the same conditions.

The first objective in this project is to work in a test bed that provides a common framework to evaluate and compare the models. This test bed presents a game that has to be played by autonomous agents using reputation models and that compete among them.

We have organised two of these competitions that have received good support from the community both in terms of participation and interest. The second objective is the advance in the research on cognitive models of reputation and how they can be integrated in the environment of an electronic institution. The improvement of the Repage model and the study and implementation of its integration in an electronic institution is the main result obtained.