隐私保护关联挖掘在职务犯罪预警中的应用

Application of privacy protection association mining in job-related crime early warning

运用数据挖掘对职务犯罪行为进行分析,揭发事实真相,并且对其他公职人员进行预警是学者们研究的热点.利用分布式环境下关联挖掘算法对数据进行分析,为了避免隐私数据的泄漏,引入安全多方计算来保护公职人员的隐私数据.通过对公职人员数据进行分类,与职务犯罪模型进行监测和对比,提前预警发现职务犯罪行为,减少职务犯罪案件的发生.利用安全多方计算中的同态加密对敏感数据进行加密保护,将不同类的数据挖掘任务转换成特定的安全多方计算问题,使数据挖掘过程更加安全可靠.

Using data mining to analyze job-related crimes,expose the truth,and give early warning to other public officials is the focus of our research.In order to avoid the leakage of private data,we introduce secure multi-party computing to protect the private data of public officials.By classifying the data of public officials,monitoring and comparing with the job-related crime model,early warning can find job-related crimes and reduce the occurrence of job-related crimes.We use homomorphic encryption in secure multi-party computing to encrypt and protect sensitive data,and transform different types of data mining tasks into specific secure multi-party computing problems.These efforts can make the data mining process more secure and reliable.