Categorization of Cyberbullying based on Intentional Dimension

Cyberbullying is a rampant issue with lasting impacts on victims' mental health and social well-being. The pervasiveness of online harassment necessitates clear definitions for effective intervention. However, research suffers from a lack of consistent terminology. A confusing array of terms describes various cyberbullying tactics, hindering clear comparisons between studies and impeding the development of targeted solutions. Further, this gap is particularly problematic for advancements in computing and machine learning applied to cyberbullying detection and prevention. This paper addresses this gap by reviewing the literature on cyberbullying terminology and definitions. The methodology employed herein was to analyze the array of terms used, highlighting the inconsistencies. This analysis leads to a novel categorization framework based on the intentional dimension of cyberbullying behaviors. We categorized cyberbullying tactics based on their core malicious intent. This categorization is critical for future research for three key reasons. First, it provides a standardized language for researchers to compare cyberbullying behaviors. Second, it provides consistent terminology for algorithms designed to identify and flag cyberbullying behavior. With clear definitions, machine learning models can differentiate between cyberbullying behaviors based on intention. Our research review and analysis lead to the cyberbullying categorization includes intentions to provoke anger (flaming), damage reputation (denigration), instill fear (threats), isolate (exclusion), expose secrets (doxing), exploit (sexting), and an umbrella term for upset, humiliate, and distress (harassment).