社交网络敏感信息识别系统的设计研究  

Design and research on sensitive information recognition system of social network

作　　者：张振清 张凯存[2] 李卫平[3] ZHANG Zhen-qing;ZHANG Kai-cun;LI Wei-ping(Department of Criminal Science and technology,Railway Police College,Zhengzhou 450053,China;Department of Public Foundation,Railway Police College,Zhengzhou 450053,China;Department of Image and Network Investigation,Railway Police College,Zhengzhou 450053,China)

机构地区：[1]铁道警察学院刑事科学技术系,郑州450053 [2]铁道警察学院基础部,郑州450053 [3]铁道警察学院图像与网络侦查系,郑州450053

出　　处：《信息技术》2021年第7期59-64,共6页Information Technology

基　　金：河南省科技攻关项目(182102210118,182102210106);公安部技术研究计划项目(2018JSYJC23);公安部公安理论及软科学研究计划重点项目(2019LLYJHNST008);河南省高等学校重点科研项目(19B520026);中央高校基本科研业务经费项目(2017TJJBKY001)。

摘　　要：文中介绍了社交网络敏感信息识别系统的技术架构和功能模块,并对系统的总体部署进行了深入探讨。对社交网络数据进行了采集、清洗、分类、分析和处理,对相关算法进行了改进和优化。实验结果表明,无论是半监督场景还是监督场景,提出的算法在主题分类和情绪分类中都有比较理想的性能,在文本分类的各个子领域具有广泛的应用价值。The paper introduces the technical architecture and functional modules of the sensitive information recognition system of social network,and discusses the overall deployment of the system.After the social network data has been collected,cleaned,classified,analyzed and processed,we improved and optimized related algorithms.The experiment results show that the algorithm proposed has better performance in subject classification and sentiment classification,whether it is a semi-supervised scenario or a supervised scenario,which has extensive application value in various subfields of text classification.

关 键 词：敏感信息 分类系统 技术架构 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构]