Modeling and Evaluating Trust Network Inference

This paper formalizes trust network inference notions, providing both data and computational models, and suggests an evaluation model for benchmarking that is more customizable, customizable, scalable and scalable than traditional approaches such as public reputation systems and collaborative collaborative approaches.

Abstract

The growth in knowledge sharing enabled by the (Semantic) Web has made trust an increasingly critical issue. Based on explicit inter-agent trust relations, a trust network emerges on the (Semantic) Web in the knowledge sharing context. The concept of a trust network and its application to knowledge sharing have received recent attention but neither their structural properties (e.g. dynamics, complexity) nor inference mechanisms (e.g. trust discovery, trust evolution, trust propagation) have been well addressed. This paper formalizes trust network inference notions, providing both data and computational models, and suggests an evaluation model for benchmarking. The data model clarifies the data (context, restriction, output) used by trust network inference for knowledge sharing. It also elaborates trust network representation and articulates different types of trust. The computational model reviews graph theory and referral network interpretations of trust network inference and proposes a new one that treats trust network as an emergent property. This new model supports both trust evolution and trust propagation. The evaluation model describes metrics as well as methods to generate test scenarios and data. We argue that this approach is more customizable, flexible and scalable than traditional approaches such as public reputation systems and collaborative filtering. 1.