The Internet of Vehicles (IoV) has emerged as a 1 new spin-off research theme from traditional vehicular ad hoc 2 networks. It employs vehicular nodes connected to other smart 3 objects equipped with a powerful multisensor platform, commu-4 nication technologies, and IP-based connectivity to the Internet, 5 thereby creating a possible social network called Social IoV 6 (SIoV). Ensuring the required trustiness among communicating 7 entities is an important task in such heterogeneous networks, 8 especially for safety-related applications. Thus, in addition to 9 securing intervehicle communication, the driver/passengers hon-10 esty factor must also be considered, since they could tamper 11 the system in order to provoke unwanted situations. To bridge 12 the gaps between these two paradigms, we envision to connect 13 SIoV and online social networks (OSNs) for the purpose of 14 estimating the drivers and passengers honesty based on their 15 OSN profiles. Furthermore, we compare the current location of 16 the vehicles with their estimated path based on their historical 17 mobility profile. We combine SIoV, path-based and OSN-based 18 trusts to compute the overall trust for different vehicles and their 19 current users. As a result, we propose a trust-aware communi-20 cation architecture for social IoV (TACASHI). TACASHI offers 21 a trust-aware social in-vehicle and intervehicle communication 22 architecture for SIoV considering also the drivers honesty factor 23 based on OSN. Extensive simulation results evidence the effi-24 ciency of our proposal, ensuring high detection ratios >87% and 25 high accuracy with reduced error ratios, clearly outperforming 26 previous proposals, known as RTM and AD-IoV. 27 Index Terms-Human factor, Social Internet of Vehicles (SIoV), 28 trust, vehicular ad hoc network (VANET). 29 I. INTRODUCTION 30 M ANY applications have been realized through vehic-31 ular networks as a result of communication among 32