A Client-Aware Reputation System for e-Services

In the Internet age, people are becoming more an more familiar in experiencing online services. Given the intrinsic distributed nature of the electronic transactions involved, there is the need to prove somehow the trustworthiness of such services for supporting a user in her choice. To this aim, different techniques have been presented. For example, a simple but useful solution is to rely on feedback of past users testifying if they have been satisfied by a service. In this paper, we consider a scenario for business transactions where a reputation management system helps clients in choosing services that best satisfy their attitudes and preferences. Specifically, a reputation value is associated to each service at stake. This value is updated according to past and new clients interactions. In fact, at the end of each interaction with a service, clients provide feedback regarding that service. The main feature of our proposal is the client awareness. This derives from designing and implementing a probabilistic client model based on real behaviours of users when they choose a service and give feedback. This client model has been obtained by collecting and processing real data from ones of the most popular websites for travel advice. We present an evaluation aimed at validating our proposal. In the simulations, we also deal with the issue of false feedback, reported by clients that intentionally aim at subverting the reputation value of a service. The simulations results show that our system is robust up to a certain number of malicious feedback.