In this article, we propose the PTP-MF (Pairwise Trust Prediction through Matrix Factorisation) algorithm, an approach to predicting the intensity of trust and distrust relations in Online Social Networks (OSNs). Our algorithm maps each OSN user i onto two low-dimensional vectors, namely, the trustor profile (describing her/his inclination to trust others) and the trustee profile (modelling how others perceive i as trustworthy) and it computes the trust a user i places in a user j as the dot product of trustor profile of i and the trustee profile of j. The PTP-MF algorithm incorporates also biases in trustor and trustee behaviour to make more accurate predictions. Experiments on four real-life datasets indicate that the PTP-MF algorithm significantly outperforms other methods in accuracy and it showcases a high scalability.
Trust Prediction via Matrix Factorisation
De Meo P.
2019-01-01
Abstract
In this article, we propose the PTP-MF (Pairwise Trust Prediction through Matrix Factorisation) algorithm, an approach to predicting the intensity of trust and distrust relations in Online Social Networks (OSNs). Our algorithm maps each OSN user i onto two low-dimensional vectors, namely, the trustor profile (describing her/his inclination to trust others) and the trustee profile (modelling how others perceive i as trustworthy) and it computes the trust a user i places in a user j as the dot product of trustor profile of i and the trustee profile of j. The PTP-MF algorithm incorporates also biases in trustor and trustee behaviour to make more accurate predictions. Experiments on four real-life datasets indicate that the PTP-MF algorithm significantly outperforms other methods in accuracy and it showcases a high scalability.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
