2026-04-17T17:44:45.931Z http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai

ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai:3:10167 2024-12-03T20:10:47Z 3

Natural Language Processing and AdaBoost Optimized by Modified Metaheuristic for Online Harassment Detection, Chapter in ISEM Information Systems Engineering and Management: ICIACS 2024: Innovations and Advances in Cognitive Systems, Springer, volume 16 2024 http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai/record/3/10167 https://link.springer.com/chapter/10.1007/978-3-031-69201-7_33 L. Jovanovic N. Bacanin B. Radomirovic M. Zivkovic A. Njegus M. Antonijevic In the realm of digital interactions, cyberbullying stands as a significant yet often overlooked concern within scholarly investigations. The presence of hostile online environments not only discourages active participation but can also affect mental well-being. Pinpointing instances of cyberbullying poses a challenge. With the increasing number of platforms facilitating user engagement through comments and feedback, traditional moderation methods prove inadequate. Moreover, the dynamic nature of online interactions necessitates adaptable strategies, as definitions of aggressive behavior continuously evolve. In light of these complexities, this study advocates for a data-driven approach. The potential of leveraging optimization metaheuristics alongside robust classification algorithms and bidirectional encoder representations from transformers encoders for detecting instances of online attacks is explored in this work. AdaBoost hyperparameters are subjected to optimization through established and emergent algorithms, with a modified version of the botox optimization algorithm introduced specifically for the needs of this study. The efficacy of this framework was evaluated using a publicly accessible dataset. The optimized model attained an accuracy rate of 93.39% suggesting the viability of the approach for confronting the escalating challenges posed by cyberbullying. bookPart Springer, Cham 446 463 10.1007/978-3-031-69201-7_33 ISEM Information Systems Engineering and Management: ICIACS 2024: proceedings of the International Conference on Innovations and Advances in Cognitive Systems, volume 16