研究人员开发工具,在网上发现变相仇恨言论,加强内容的节制。 Researchers develop tool to detect disguised hate speech online, enhancing content moderation.
奥克兰大学的研究人员创造了一个工具,通过发现变相的仇恨评论来改进网上内容的调和。 Researchers at the University of Auckland have created a tool to improve online content moderation by detecting disguised hate comments. 该工具确定了用数字取代字母或改变字词等战术,并帮助传统过滤器发现隐藏的毒性。 The tool identifies tactics like replacing letters with numbers or altering words and helps traditional filters spot hidden toxicity. 这项进展旨在保护用户,特别是年轻受众,并通过加强对有害内容的检测,确保更安全的在线环境。 This advancement aims to protect users, especially younger audiences, and ensure safer online environments by enhancing the detection of harmful content. 今后的发展可能包括更深入的背景分析。 Future developments may include deeper context analysis.