Trusted Friends’ Computation Method Considering Social Network Interactions’ Time

Abstract

Based on the assumption that users generally tend to use entities proposed by friends rather than strangers and that trust among friends significantly correlates with user’s trends, we decided to refer to research conducted on the evolving field of social trust computation. Although many models were proposed to analyze computational trust for various applications of social web, little importance is given to the time factor even by models that represent trust as a source of recommendation measurement. In this paper, we propose to integrate the temporal factor in assessing trust between social network friends. We define a time-aware method for an implicit trust determination between Facebook friends. Then, we integrate it in a mobile app we have developed for identifying trusted friends in social networks. Finally, we assess our method using an objective statistical method and compare it with others methods in the literature to show the importance of the time of interactions between users for a better calculation of social trust and consequently for an enhanced identification of trusted friends in social networks.