俄语虚假新闻文体计量研究  

A Corpus-based Study on the Measurement of the Stylistic Features of Russian Fake News

作　　者：原伟 刘海涛[2] Yuan Wei;Liu Hai-tao(College of Foreign Studies,National University of Defense Technology,Nanjing 210039,China;School of International Studies,Zhejiang University,Hangzhou 310058,China)

机构地区：[1]国防科技大学外国语学院,南京210039 [2]浙江大学外国语学院,杭州310058

出　　处：《外语学刊》2024年第6期1-8,共8页Foreign Language Research

基　　金：国家社科基金项目“生成式人工智能赋能语言学多模态学科知识图谱建设与应用研究”(24BYY079);河南省哲学社会科学规划项目“基于外媒多模态新闻可比语料库的涉华突发事件监测与态势感知”(2021BYY024)的阶段性成果。

摘　　要：俄语虚假新闻的数量猛增且影响范围扩大,其分析与自动检验问题值得关注。本文基于语料库采用计量文体学方法考察俄语虚假新闻在词、句和情感表达上的特点,结果发现:虚假新闻词汇丰富度较高,但信息量和难度偏低,更倾向于使用多句、短句、结构简单的句子传递信息;14个词计量特征、26个句计量特征和5个情感计量特征可用于真假新闻的自动区分聚类。该结果对于虚假新闻的文本计量研究、其他语种的虚假新闻分析、检验和防范具有启示意义。The number of Russian fake news increases significantly,and analyzing and automatic checking them deserves attention.We investigated the characteristics of Russian fake news in terms of words,sentences,and emotions using a stylometric approach,and found that:fake news has a high lexical richness but low information content and difficulty,and tends to use multiple sentences,short sentences,and simple structure to convey information;14 word metric features,26 sentence metric features,and 5 emotion metric features can be effectively used to automatically differentiate real and fake news clustering.The results have implications for textual measurement of fake news and analyzing,investigating,and preventing fake news in other languages.

关 键 词：俄语 虚假新闻 文体计量 语料库 

分 类 号：H030[语言文字—语言学]