A novel iterative online rating attack based on market self-exciting property

摘要

The prosperity of online rating system makes it an important place for malicious vendors to mislead public's online decisions, whereas the security related studies are lagging behind. In this work, we adopt a quantile regression model to investigate influential factors on online user choices and reveal the "self-exciting" property of online market. Inspired by these findings, we propose a novel iterative rating attack and validate its advantage through experiments.