基于K近邻分类算法的网络敏感信息自动过滤  

Automatic Filtering of Network Sensitive Information Based on K Nearest Neighbor Classification Algorithm

作　　者：石小兵[1] SHI Xiao-bing(Anhui Vocational College of Industrial Economics,Hefei,Anhui 230001,China)

机构地区：[1]安徽工业经济职业技术学院,安徽合肥230001

出　　处：《河北北方学院学报（自然科学版）》2021年第11期1-6,共6页Journal of Hebei North University:Natural Science Edition

摘　　要：目的为了提高网络敏感信息过滤能力,提出基于K近邻分类算法的网络敏感信息自动过滤方法。方法采用混合云构架技术对网络敏感信息云存储结构进行分析,根据敏感信息结构特征建立信息特征融合和空间特征压缩模型,采用自相关特征匹配方法实现对网络敏感信息滤波和特征点标定,通过模糊度检测和云融合技术实现对信息的融合处理,采用K近邻分类算法构建网络敏感信息聚类和网格分块重组模型,实现对网络敏感信息自动过滤。结果仿真结果表明,该方法的网络敏感信息检测精度始终保持在90%以上,检测精度较高,且网络敏感信息过滤的吞吐量较高。结论可以有效提升网络敏感信息过滤能力,实际应用效果好。Objective To improve the filtering ability of network sensitive information,an automatic filtering method of network sensitive information based on K-nearest neighbor classification algorithm was proposed.Methods The hybrid cloud architecture technology was used to analyze the cloud storage structure of network sensitive information.According to the structure characteristics of sensitive information,the information feature fusion and spatial feature compression model were established.The autocorrelation feature matching method was used to realize the filtering and feature point calibration of network sensitive information.The information fusion processing was realized through the ambiguity detection and cloud fusion technology,and the K-nearest neighbor classification algorithm was used to construct the cloud storage structure.The model of clustering and grid partition reorganization of network sensitive information was built to realize the automatic filtering of network sensitive information.Results The simulation results showed that the detection accuracy of this method was always above 90%,and the throughput of network sensitive information filtering was high.Conclusions This method can effectively improve the filtering ability of network sensitive information,and it has good practical application effect.

关 键 词：K近邻分类 网络敏感信息 自动过滤 自相关特征匹配方法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]