Detecting and Analyzing the Malicious Social Bots by using Data Mining and Naïve Bayesian Classifier

Abstract

With the huge increase in the number, speed, and variety of customer information (such as customer-generated information) in online interpersonal organizations, people have worked hard to construct better methods to collect and inspect such large information. For example, social robots have been used to perform scientific management of robots and provide customers with improved management attributes. Even so, harmful social robots are used to spread false data (for example, false news), which may bring real results. Therefore, identifying and evacuating harmful social robots in online informal communities is crucial. The latest discovery technology of resentful social robots undermines the quantitative focus of its behavior. These highlights are easily imitated by social robots. This leads to a reduction in the accuracy of the investigation. This article describes an epic strategy for identifying malicious social robots, which includes two key options, depending on the possibility of snapshot stream grouping and semi-hosted binding changes. The technology not only investigates the possibility of changes in the click stream of customer behavior, but also takes into account the temporal highlights of the behavior.