Peer Prediction-Based Trustworthiness Evaluation and Trustworthy Service Rating in Social Networks

With the development of online applications based on the social network, many different approaches of service to achieve these applications have emerged. Users’ reporting and sharing of their consumption experience or opinion can be utilized to rate the quality of different approaches of online services. How to ensure the authenticity of the users’ reports and identify malicious ones with cheating reports become important issues to achieve an accurate service rating. In this paper, we provide a private-prior peer prediction mechanism based trustworthy service rating system with a data processing center (DPC), which requires users to report to it with their prior and posterior believes that their peer users will report a high quality opinion of the service. The DPC evaluates users’ trustworthiness with their reports by applying the strictly proper scoring rule, and removes reports received from users with low trustworthiness from the service rating procedure. This peer prediction method is incentive compatible and able to motivate users to report honestly. In addition, to identify malicious users and bad-functioning/unreliable users with high error rate of quality judgement, an unreliability index is proposed in this paper to evaluate the uncertainty of reports. Reports with high unreliability values will also be excluded from the service rating system. By combining the trustworthiness and unreliability, malicious users will face a dilemma that they cannot receive a high trustworthiness and low unreliability at the same time when they report falsely. Simulation results indicate that the proposed peer prediction based trustworthy service rating can identify malicious and unreliable behaviours effectively, and motivate users to report truthfully. The relatively high service rating accuracy can be achieved by the proposed system