网络安全与数据治理2025，Vol.44Issue(4)：79-83,5.DOI:10.19358/j.issn.2097-1788.2025.04.012
藏文网络敏感信息检测研究
Research on sensitive information detection in Tibetan network
摘要
Abstract
With the increasing popularity of the Internet,the Tibetan-language online space is facing growing risks of sensitive in-formation dissemination,posing challenges to social stability and national security.Traditional methods for detecting sensitive in-formation are unable to effectively address the unique characteristics of the Tibetan language and the complexity of online informa-tion.To address this issue,this paper proposes a hybrid neural network model based on CINO-DPCNN.This model combines the deep understanding of Tibetan semantics provided by the CINO model with the high-efficiency feature extraction capabilities of the DPCNN model,enabling more accurate identification of sensitive information in Tibetan-language online networks.The experimental results demonstrate that the CINO-DPCNN model has achieved excellent performance in terms of accuracy,F1 score,and other indi-cators,showing significant improvements over existing models.This provides new technical support for building a secure and healthy Tibetan-language online environment and serves as a reference for sensitive information detection in other minority languages.关键词
藏文/敏感信息/CINO模型/深度学习Key words
Tibetan/sensitive information/CINO model/deep learning分类
计算机与自动化引用本文复制引用
吴瑜,严李强,徐梓恒,卓玛央金..藏文网络敏感信息检测研究[J].网络安全与数据治理,2025,44(4):79-83,5.基金项目
国家自然科学基金项目(62406256) （62406256）
西藏大学研究生高水平人才培养计划项目(2022-GSP-S105) （2022-GSP-S105）
