Using Mobile Nodes to Control Rumors in Big Data Based on a New Rumor Propagation Model in Vehicular Social Networks

Abstract

Vehicle data, which may have some errors, provide a source for big data analysis. Car owners sometimes publish false information to protect their privacy or interests. The spread of these false messages will contribute to potential loss as rumors. Recently, the emergence of the vehicular social network has further accelerated the spread of rumors and anti-rumors. Researchers have proposed a number of methods to reduce the loss caused by rumors. However, these methods do not account for the fact that it takes some time for users to reply to the message after receiving the rumors. Each user responds differently to rumors, but previous methods do use fixed conversion rates in different states. To overcome these difficulties, individuals are classified as susceptible, trustful, contagious, immune or recoverable (STCIR). We propose a novel STCIR model to study the dynamic propagation of rumors. Each user is assigned a time threshold attribute to indicate the time delay of the user's response. In some cases, a user may receive a rumor repeatedly in the course of the time threshold. Furthermore, we introduce vehicular nodes as authorities that produce correct information to curb rumors through user forwarding, and the anti-rumors starting from users are regarded as a rumor cascade. The experimental results reveal that vehicular nodes can reduce the scale and number of rumors. Vehicular nodes emerge early and move fast, and by moving to a greater degree, these nodes can increase the efficiency of mitigating rumors.