摘要
以深度学习发展为线索,从域角度对假新闻检测模型进行分类,讨论虚假新闻检测方法的变化。首先在概述中分别对域角度下虚假新闻检测的定义、域适用数据集、假新闻检测评估指标及相关工作进行介绍;接着从域角度将虚假新闻检测模型归纳总结成3类,分别为源域与目标域相同、源域与目标域不同和忽略域信息3种类型的检测模型,并对3类模型以深度学习为线索进行梳理。
This paper follows the development of deep learning to classify fake news detection models from a domain perspective and examineses the evolution of detection methods First,the overview introduces the definition of fake news detection from a domain perspective,domain-specific datasets,evaluation metrics for fake news detection,and related work.Then,based on the domain perspective,fake news detection models are categorized into three types:those with the same source and target domains,those with different source and target domains,and those ignoring domain information.These three types of models are systematically reviewed with a focus on deep learning.
作者
陈烁淳
黄发良
戴智鹏
黄恩博
CHEN Shuochun;HUANG Faliang;DAI Zhipeng;HUANG Enbo(Guangxi Key Lab of Human-machine Interaction and Intelligent Decision,Nanning Normal University,Nanning 530199,China)
出处
《福建师范大学学报(自然科学版)》
北大核心
2025年第2期43-54,116,共13页
Journal of Fujian Normal University:Natural Science Edition
基金
国家自然科学基金项目(62262045)
广西重点研发计划项目(桂科AB22035072)。
关键词
虚假新闻检测
深度学习
域视角
人工智能
fake news detection
deep learning
domain perspective
artificial intelligence
