Identifying Causal Structures from Cyberstalking: Behaviors Severity and Association

Abstract

This paper presents an etiological cyberstalking study, meaning the use of various technologies and internet in general to harass or to stalk someone. The novelty of the paper is the multivariate empirical approach of cyberstalking victimization that has received less attention from the research community. Also, there is a lack of such studies from the causal perspective. It happens, since in most of the studies, a priority is given on a single causation identification, whereas the data examination used for mining causal relationships in this paper presents a novel and great potential to detect combined or multiple cause factors. The paper focuses in the impact that variables such as age, gender and the fact whether the participant has ever harassed someone, is related to the fact of being victim of cyberstalking. The research aims to find the causes of cyberstalking in high school’s teenagers. Furthermore, an exploratory data analysis has been performed. A weak and moderate correlation between the factors on the dataset is emphasized. The odds ratio among the variables has been calculated, which implies that girls are twice as likely as boys to be cyberstalked. Similarly, concerning outcomes related to cyberstalking frequency recidivism are noticed.

Keywords

• causality
• causal rules
• cyberstalking
• data mining