通信学报2026，Vol.47Issue(4)：113-125,13.DOI:10.11959/j.issn.1000-436x.2026083
基于增强关系图卷积网络的数据违规转售检测方法
Illicit data resale detection method via an enhanced relational graph convolutional network
摘要
Abstract
Illicit data resale in data trading scenarios exhibited strong concealment and was difficult to detect.An en-hanced relational graph convolutional network was proposed by optimizing message passing and feature aggregation with transaction contextual similarity and causal temporal order constraints,enabling effective representation of illicit re-sale behaviors under complex transaction relations.Based on this model,a detection method was developed to predict the existence of illicit resale behaviors in transaction topology graphs.A simulated data trading dataset containing anoma-lous resale samples was constructed,and comparative experiments were performed.The results indicate that the pro-posed method provides an effective solution for illicit data resale detection in data trading scenarios.关键词
数据流通交易/数据违规转售/关系图卷积网络/注意力机制Key words
data trading/illicit data resale/relational graph convolutional network/attention mechanism分类
信息技术与安全科学引用本文复制引用
王宇翔,张玲翠,侯雨桥,杨倩,牛犇..基于增强关系图卷积网络的数据违规转售检测方法[J].通信学报,2026,47(4):113-125,13.基金项目
国家重点研发计划基金资助项目(No.2023YFB3106505) （No.2023YFB3106505）
国家自然科学基金资助项目(No.U24A20240,No.62441226) The National Key Research and Development Program of China(No.2023YFB3106505),The National Natural Science Foundation of China(No.U24A20240,No.62441226) （No.U24A20240,No.62441226）
