Hate Cannot Drive Out Hate: Forecasting Conversation Incivility following Replies to Hate Speech
Uložené v:
| Názov: | Hate Cannot Drive Out Hate: Forecasting Conversation Incivility following Replies to Hate Speech |
|---|---|
| Autori: | Yu, Xinchen, Blanco, Eduardo, Hong, Lingzi |
| Zdroj: | Proceedings of the International AAAI Conference on Web and Social Media. 18:1740-1752 |
| Publication Status: | Preprint |
| Informácie o vydavateľovi: | Association for the Advancement of Artificial Intelligence (AAAI), 2024. |
| Rok vydania: | 2024 |
| Predmety: | FOS: Computer and information sciences, Computer Science - Computers and Society, 03 medical and health sciences, Computer Science - Computation and Language, 0302 clinical medicine, 0504 sociology, 05 social sciences, Computers and Society (cs.CY), Computation and Language (cs.CL) |
| Popis: | User-generated counter hate speech is a promising means to combat hate speech, but questions about whether it can stop incivility in follow-up conversations linger. We argue that effective counter hate speech stops incivility from emerging in follow-up conversations—counter hate that elicits more incivility is counterproductive. This study introduces the task of predicting the incivility of conversations following replies to hate speech. We first propose a metric to measure conversation incivility based on the number of civil and uncivil comments as well as the unique authors involved in the discourse. Our metric approximates human judgments more accurately than previous metrics. We then use the metric to evaluate the outcomes of replies to hate speech. A linguistic analysis uncovers the differences in the language of replies that elicit follow-up conversations with high and low incivility. Experimental results show that forecasting incivility is challenging. We close with a qualitative analysis shedding light into the most common errors made by the best model. |
| Druh dokumentu: | Article |
| ISSN: | 2334-0770 2162-3449 |
| DOI: | 10.1609/icwsm.v18i1.31422 |
| DOI: | 10.48550/arxiv.2312.04804 |
| Prístupová URL adresa: | http://arxiv.org/abs/2312.04804 |
| Rights: | arXiv Non-Exclusive Distribution |
| Prístupové číslo: | edsair.doi.dedup.....0fff8b8294016191e51f3d96044ac64a |
| Databáza: | OpenAIRE |
Nájsť tento článok vo Web of Science | Abstrakt: | User-generated counter hate speech is a promising means to combat hate speech, but questions about whether it can stop incivility in follow-up conversations linger. We argue that effective counter hate speech stops incivility from emerging in follow-up conversations—counter hate that elicits more incivility is counterproductive. This study introduces the task of predicting the incivility of conversations following replies to hate speech. We first propose a metric to measure conversation incivility based on the number of civil and uncivil comments as well as the unique authors involved in the discourse. Our metric approximates human judgments more accurately than previous metrics. We then use the metric to evaluate the outcomes of replies to hate speech. A linguistic analysis uncovers the differences in the language of replies that elicit follow-up conversations with high and low incivility. Experimental results show that forecasting incivility is challenging. We close with a qualitative analysis shedding light into the most common errors made by the best model. |
|---|---|
| ISSN: | 23340770 21623449 |
| DOI: | 10.1609/icwsm.v18i1.31422 |