抵御恶意子集确权攻击的数据确权方案

打开文本图片集

关键词：数据确权；恶意子集攻击；原子特征；集合包含关系；规范形式；K-means聚类；布隆过滤器中图分类号：TP391 文献标志码：A 文章编号：1001-3695（2026）04-029-1208-06doi：10.19734/j. issn.1001-3695.2025.07.0285

Data rights confirmation schemes to defend against malicious subset rights confirmation attacks

ChenQingru，Zhou Quan，Ye Xinqi （School ofMathematicsand Information Science，Guangzhou University，Guangzhou 51oo6，China）

Abstract：This studyaddressedtheownershipdisputes caused bysecondaryrightsconfirmation indata element marketization， whichhindersdata circulationeficiency.Theresearch developedastructureddatarightsconfirmationmodel todefendagainst malicioussubsetrightsconfirmationattacks.Themethoddecomposeddataintoatomicfeatureunits：categoricalatributeswere transformed intosemanticallyuniqueorthogonal vectorsviaone-hotencoding，whilenumerical atributes wereprocessed throughK-means clustering togenerateminimal semanticclustercenter vectors.Itdesignedacanonical formfunctionto eliminateorderingambiguityin featuresets，anddeterminedsubset ownershipbasedonsetinclusionrelationships，optimized thedetectionofsecondaryrightsconfirmationapplicationsusingcolision-resistanthashand Bloomfilter.Experimentalresults demonstratethatthemodelimprovesteeficiencyofdetectingilegalsubsetrightsconfirmation，withanaveragerejectionrate of 71. 45% and verification time under 35ms ， enabling feature-level fine-grained ownership control. The study concludes that theproposedschemeprovidesacomputableandverifiable solutiontomitigatesecondaryrightsconfirmationrisksindatacirculation，offering high practicality and security.

Keywords：datarightsconfirmation；malicious subsetatack；atomicfeature；set inclusionrelationship；canonical form function；K-means clustering；Bloom filter

0 引言

在数字经济加速融合实体经济的背景下，数据要素的市场化配置已成为全球竞争的战略焦点[1]。（剩余16342字）