摘要
常规的敏感图像数据挖掘方法是利用原有数据库进行图像敏感信息识别,这使新兴敏感数据无法被有效挖掘。为此,设计了基于差分隐私的网络图像敏感数据挖掘方法。首先,挖掘出图像敏感数据差分特征,剔除疑似敏感的图像数据。然后,利用差分隐私算法,生成网络图像的视觉敏感词典,有效识别新兴敏感数据。最后,构建网络图像敏感数据挖掘模型,进一步提高敏感数据挖掘精准度。实验结果表明所提方法挖掘的敏感数据更加准确。
Conventional sensitive image data mining methods use the original database to identify image sensitive information,which prevents the emerging sensitive data from being effectively mined.Therefore,a method of network image sensitive data mining based on differential privacy is designed.First,the difference features of image sensitive data are mined,and the suspected sensitive image data is eliminated.Then,using differential privacy algorithm,a visual sensitive dictionary of network images is generated to effectively identify emerging sensitive data.Finally,a sensitive data mining model of network images is constructed to further improve the accuracy of sensitive data mining,so as to achieve efficient mining of sensitive data of network images.The results show that this method is more accurate in mining sensitive data,and has higher promotion value.
作者
杜玉昌
DU Yuchang(Admissions Office,Xiamen Software Vocational and Technical College,Xiamen 361024,China)
出处
《新乡学院学报》
2023年第3期30-33,共4页
Journal of Xinxiang University
关键词
差分隐私
网络图像
敏感数据
挖掘方法
差分特征
敏感词典
differential privacy
network image
sensitive data
excavation method
differential characteristics
sensitive dictionary
