Assessing the Surge in COVID-19-Related Cyberbullying on Twitter: A Generalized Additive Model Approach
Abstract
The COVID-19 pandemic's onset and the subsequent lockdowns drastically amplified digital interactions worldwide. These unparalleled shifts in online behavior birthed concerns about potential surges in cybersecurity threats, particularly cyberbullying. Our research aimed to explore these proposed trends on Twitter. Utilizing a dataset of 126,348 tweets from January 1st to September 12th, 2020, we honed in on 27 cyberbullying-related keywords, like 'online bullying' and 'cyberbullying'. Recognizing the limitations of traditional change-point models, we opted for a Generalized Additive Model (GAM) with spline-based smoothers. The results were revealing. A significant uptick in cyberbullying instances emerged starting mid-March, correlating with the global lockdown mandates. This consistent trend was evident across all our targeted keywords. To bolster our findings, we conducted lag-based assessments and compared the GAM against other modeling approaches. Our conclusions robustly indicate a strong association between the enforcement of pandemic lockdowns and a heightened prevalence of cyberbullying on Twitter. The implications are clear: global crises necessitate intensified cyber vigilance, and the digital realm's safety becomes even more paramount during such challenging times.
Keywords
covid19 , Cyberbullying , Twitter , GAM
References
- Achuthan, K., Nair, V. K., Kowalski, R., Ramanathan, S., & Raman, R. (2023). Cyberbullying research — Alignment to sustainable development and impact of COVID-19: Bibliometrics and science mapping analysis. Computers in Human Behavior, 140, 107566. https://doi.org/10.1016/j.chb.2022.107566
- Balakrishnan, V., Khan, S., & Arabnia, H. R. (2020). Improving cyberbullying detection using Twitter users’ psychological features and machine learning. Computers & Security, 90, 101710. https://doi.org/10.1016/j.cose.2019.101710
- Brandt, P. T., & Sandler, T. (2012). A Bayesian Poisson vector autoregression model. Political Analysis, 20(3), 292-315.
- Bonanno, R. A., & Hymel, S. (2013). Cyber Bullying and Internalizing Difficulties: Above and Beyond the Impact of Traditional Forms of Bullying. Journal of Youth and Adolescence, 42(5), 685-697. https://doi.org/10.1007/s10964-013-9937-1
- Cerna, M. A. (2015). The Chinese “Togetherness-Separation” Paradox: An Analytical Approach to Understanding Chinese People’s Behavior and Its Implication to International Cooperation. International Journal of Business and Management, 10(12), 194. https://doi.org/10.5539/ijbm.v10n12p194
- Chelmis, C., Zois, D.-S., & Yao, M. (2017). Mining Patterns of Cyberbullying on Twitter. 2017 IEEE International Conference on Data Mining Workshops (ICDMW), 126-133. https://doi.org/10.1109/ICDMW.2017.22
- Cheng, L., Shu, K., Wu, S., Silva, Y. N., Hall, D. L., & Liu, H. (2020). Unsupervised Cyberbullying Detection via Time-Informed Gaussian Mixture Model (arXiv:2008.02642). arXiv. https://doi.org/10.48550/arXiv.2008.02642
- Cuadrado-Gordillo, I., & Fernández-Antelo, I. (2016). Vulnerability and Mimicry as Predictive Axes in Cyberbullying. Journal of Interpersonal Violence, 31(1), 81-99. https://doi.org/10.1177/0886260514555128
- Das, S., Kim, A., & Karmakar, S. (2020). Change-point analysis of cyberbullying-related twitter discussions during covid-19. arXiv preprint arXiv:2008.13613.
- Dewani, A., Memon, M. A., & Bhatti, S. (2021). Development of computational linguistic resources for automated detection of textual cyberbullying threats in Roman Urdu language. 3 c TIC: Cuadernos de Desarrollo Aplicados a Las TIC, 10(2), 101-121.