Year: 2023
Type: conference-paper
Abstract: This paper describes a system for monitoring in real time the level of violence on Web platforms through the use of an artificial intelligence model to classify textual data according to their content. The system was successfully implemented and tested during the electoral campaign period of the Brazilian 2022 elections by using it to monitor the attacks directed to thousands of candidates on Twitter. We show that, despite an accurate and absolute quantification of violence is not feasible, the system yields differential measures of violence levels that can be useful for understanding human behavior online.
Author: Henrique S. Xavier
Institution: Brazilian Network Information Center
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4367309755