SynBullying: A Multi LLM Synthetic Conversational Dataset for Cyberbullying Detection
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915875037118464 |
|---|---|
| author | Kazemi, Arefeh Qadeer, Hamza Wagner, Joachim Hosseini, Hossein Kalaivendan, Sri Balaaji Natarajan Davis, Brian |
| author_facet | Kazemi, Arefeh Qadeer, Hamza Wagner, Joachim Hosseini, Hossein Kalaivendan, Sri Balaaji Natarajan Davis, Brian |
| contents | We introduce SynBullying, a synthetic multi-LLM conversational dataset for studying and detecting cyberbullying (CB). SynBullying provides a scalable and ethically safe alternative to human data collection by leveraging large language models (LLMs) to simulate realistic bullying interactions. The dataset offers (i) conversational structure, capturing multi-turn exchanges rather than isolated posts; (ii) context-aware annotations, where harmfulness is assessed within the conversational flow considering context, intent, and discourse dynamics; and (iii) fine-grained labeling, covering various CB categories for detailed linguistic and behavioral analysis. We evaluate SynBullying across five dimensions, including conversational structure, lexical patterns, sentiment/toxicity, role dynamics, harm intensity, and CB-type distribution. We further examine its utility by testing its performance as standalone training data and as an augmentation source for CB classification. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_11599 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | SynBullying: A Multi LLM Synthetic Conversational Dataset for Cyberbullying Detection Kazemi, Arefeh Qadeer, Hamza Wagner, Joachim Hosseini, Hossein Kalaivendan, Sri Balaaji Natarajan Davis, Brian Artificial Intelligence Computation and Language Computers and Society We introduce SynBullying, a synthetic multi-LLM conversational dataset for studying and detecting cyberbullying (CB). SynBullying provides a scalable and ethically safe alternative to human data collection by leveraging large language models (LLMs) to simulate realistic bullying interactions. The dataset offers (i) conversational structure, capturing multi-turn exchanges rather than isolated posts; (ii) context-aware annotations, where harmfulness is assessed within the conversational flow considering context, intent, and discourse dynamics; and (iii) fine-grained labeling, covering various CB categories for detailed linguistic and behavioral analysis. We evaluate SynBullying across five dimensions, including conversational structure, lexical patterns, sentiment/toxicity, role dynamics, harm intensity, and CB-type distribution. We further examine its utility by testing its performance as standalone training data and as an augmentation source for CB classification. |
| title | SynBullying: A Multi LLM Synthetic Conversational Dataset for Cyberbullying Detection |
| topic | Artificial Intelligence Computation and Language Computers and Society |
| url | https://arxiv.org/abs/2511.11599 |