| 苏林茂 宋学贵 向 一 王振磊.大数据与人工智能融合下敏感数据识别与脱敏技术在网络监管中的应用[J].中国科技资源导刊,2025,(2):27~39 |
| 大数据与人工智能融合下敏感数据识别与脱敏技术在网络监管中的应用 |
| Application of Sensitive Data Identification and Desensitization Technology in Network Supervision under the Integration of Big Data and Artificial Intelligence |
| 投稿时间：2024-12-04 |
| DOI： |
| 中文关键词: 脱敏技术;敏感数据识别;网络监管;安全监测;大数据;人工智能 |
| 英文关键词: desensitization technology, sensitive data identification, network supervision, security monitoring,big data, artificial intelligence |
| 基金项目: |
| 作者 | 单位 | | 苏林茂 宋学贵 向 一 王振磊 | (中国人民解放军 31656 部队，四川乐山 614221) |
|
| 摘要点击次数: 14 |
| 全文下载次数: 21 |
| 中文摘要: |
| 首先，在探讨在大数据与人工智能融合背景下，敏感数据识别与脱敏技术在网络监管中应用的基础上，针对传统数据监测方法存在的自动化程度低、监测精度不足及覆盖范围小等问题，提出一种基于敏感数据识别与脱敏技术的安全监测方法。此方法通过无监督训练和非结构化识别算法，结合大数据和人工智能技术，实现对敏感数据的自动检测和识别。实验结果显示，用户敏感行为预测准确率为94%，敏感数据解析、流量监测、页面捕获及数据采集的总体准确率为93%。然后，阐述数据预处理、识别算法及算法效果评估等步骤，采用散列函数法对识别出的敏感数据进行加密或替换，确保数据隐私性和安全性，并构建涵盖数据采集、数据共享、数据传输及数据处理等多个环节的安全监测模型，通过优化敏感数据扫描方法，提高了识别精度和效率。最后，分析当前面临的挑战，提出优化算法设计、加强数据预处理及探索新脱敏技术等改进建议，并展望多领域融合与拓展的应用前景。 |
| 英文摘要: |
| Firstly, based on the exploration of the application of sensitive data identification and desensitization technology in network supervision under the background of the integration of big data and artificial intelligence, a security monitoring method based on sensitive data identification and desensitization technology is proposed in response to the problems of low automation, insufficient monitoring accuracy and small coverage of traditional data monitoring methods. This method realizes the automatic detection and identification of sensitive data through unsupervised training and unstructured recognition algorithms, combined with big data and artificial intelligence technologies. Experimental results show that the prediction accuracy of user sensitive behavior is 94%, and the overall accuracy of sensitive data parsing, traffic monitoring, page capture and data collection is 93%. Then, the steps of data preprocessing, recognition algorithm and algorithm effect evaluation are expounded. The identified sensitive data is encrypted or replaced by using the hash function method to ensure data privacy and security. A security monitoring model covering multiple links such as data collection, data sharing, data transmission and data processing is constructed. By optimizing the sensitive data scanning method, the recognition accuracy and efficiency are improved. Finally, the current challenges are analyzed, and suggestions for improvement such as optimizing algorithm design, strengthening data preprocessing and exploring new desensitization technologies are proposed. The application prospects of multi-field integration and expansion are also prospected. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
