摘要
为了避免互联网上不良敏感信息的泛滥,创建干净、文明的用网环境,本文研究中文文本的敏感信息识别问题。基于调研分析,提出由敏感词库构建、可疑文本发现和敏感信息识别三阶段组成的识别框架,并给出每阶段的执行策略和方法。对基于Word2vec的敏感词库扩充方法进行了实验,结果表明该方法具有显著效果。
To prevent the proliferation of inappropriate and sensitive information on the internet and to create a clean and civilized online environment,this article investigates the issue of sensitive information detection in Chinese text.Based on survey analysis,a detection framework composed of three stages-the construction of a sensitive word library,the discovery of suspicious text,and sensitive information detection-is proposed,along with strategies and methods for each stage.Experiments were conducted on a method of expanding the sensitive word library based on Word2vec,and the results showed that this method had significant effects.
作者
董思源
王子扬
章坤
孙美凤
DONG Siyuan;WANG Ziyang;ZHANG Kun;SUN Meifeng(Guangling Collage of Yangzhou University,Yangzhou Jiangsu 225000;Yangzhou Baoyang Digital Technology Company,Yangzhou Jiangsu 225000)
出处
《软件》
2024年第3期51-53,共3页
Software
基金
2023年江苏省大学生创新创业训练计划资助项目(202313987020Y)。