Treextrust: topic-aware computational trust based on interaction experience, reputation of users with similarity and path algebra of graph in social networks

Abstract

The trust measure is the confidence or reliability among users or peers, which has been studied widely in online social networks. Most trust models are currently based on the concepts of interaction trust and reputation trust; however, various forms of interactions and analyses of the interaction contexts have not been considered fully for trust estimation. Moreover, the mechanism for computing reputation trust based on propagation lacks a clear foundation and is expensive in computation. The purpose of this paper is to present a family of computational trust models (called TreeXTrust) to estimate the trust degree of a user truster on another user trustee. Our model is a mathematical formulation that is based on an aggregation of topic-aware experience trust with various forms of interactions and topic-aware reputation trust with users’ similarity and operators on path algebra in a graph. We conducted experiments to evaluate the impacts of interaction forms and users’ interests on experience trust and the correlation of experience trust and reputation trust on overall trust estimation. Our experimental results demonstrated the following: (i) interest degrees influenced experience trust more than interaction ones did; (ii) a community’s evaluation of some trustee affected an overall trust estimation more than a truster’s individual evaluation did. Our family of models outperformed the state-of-the-art methods that have been presented in the literature and is a framework for selecting and implementing a suitable model of computational trust for our problem at hand.