A robust ensemble-based trust and reputation system against different types of intruder attacks

The trust and reputation system (TRS) has been widely used to measure trust relations among objects in online environment where it is hard to have direct experience. As the TRS became popular, TRS-attacking intruders started to appear to get unfair profits. Under these circumstances, it has become important to detect and handle the intruders in order to provide reliable online services. To solve this problem, various methods have been proposed. However, they were properly operated in the assumed environment only. In order to overcome this limitation, this study proposes an ensemble combination-based TRS which uses the ensemble combination and fuzzy theory for bonding various detectors and deriving robust-TR relations. The method proposed in this study is advantageous in that it can respond to various types of attacks in a robust manner and easily cope with a new attack or problem with decent flexibility and scalability.