基于总体失真与账户重用的协调操纵原则性检测
Principled Detection of Coordinated Manipulation from Aggregate Distortion and Account Reuse
浏览论文内容
中文总结 AI 辅助
本文提出一种以总体失真为证据的协调操纵检测方法，通过扣除匹配零期望获得带符号证据并跨账户累积，在亚马逊评论数据上显著提升ROC-AUC，并推导出精确的线性重用定律。
中文摘要 AI 辅助
协调操纵是集体性的：看似合理的账户可以共同扭曲评分、排名和参与度。现有防御主要从身份、图结构、内容或共同活动中构建证据。我们引入了一个以总体为先的证据层，将上下文级结果分布的失真视为主要证据对象。该引擎仅观察直方图、计数、分辨率和参考分布；在区间证据固定之前，身份信息被扣留。由于原始差异具有正的有限样本期望，我们减去匹配的零期望以获得带符号证据，并考虑参考不确定性。参与日志随后跨账户累积这些固定增量。我们刻画了匹配暴露散度，限制了自我影响，建立了有限时域分离，并推导了配对上下文的精确线性重用定律。我们通过受控旋转实验和对历史亚马逊评论流的配对反事实干预来评估该机制。历史评论提供行为背景；合成身份提供已知的联盟成员身份，而精确的干净孪生提供反事实对照。在对历史非捐赠者对照账户的固定攻击扫描中，将相同的操纵事件跨身份重新分配并增加重用，将账户得分ROC-AUC从0.500提升至0.797。在活动和暴露匹配的干净孪生中，频率处于随机水平，而反事实归因达到ROC-AUC 0.744。在均值保持形状干预下，Wasserstein-1和Jensen-Shannon证据分别达到ROC-AUC 0.909和0.967，而频率和基于均值的归因仍处于随机水平。总体证据补充了重复的共同活动，通过简单的未训练组合将混合机制ROC-AUC从0.750提升至0.874。
英文摘要
Coordinated manipulation is collective: plausible accounts can jointly distort ratings, rankings, and engagement. Existing defenses primarily construct evidence from identities, graphs, content, or co-activity. We introduce an aggregate-first evidence layer that treats distortion of a context-level outcome distribution as the primary evidence object. The engine observes only a histogram, count, resolution, and reference distribution; identities are withheld until interval evidence is fixed. Because raw discrepancies have positive finite-sample expectation, we subtract a matched null expectation to obtain signed evidence and account for reference uncertainty. Participation logs then accumulate these fixed increments across accounts. We characterize matched-exposure divergence, bound self-influence, establish finite-horizon separation, and derive an exact linear reuse law for paired contexts. We evaluate the mechanism with controlled rotation experiments and paired counterfactual interventions on historical Amazon review streams. Historical reviews provide the behavioral background; synthetic identities provide known coalition membership, and exact clean twins provide counterfactual controls. In a fixed-attack sweep against historical non-donor comparison accounts, reassigning the same manipulated events across identities with increasing reuse raises account-score ROC-AUC from 0.500 to 0.797. With activity- and exposure-matched clean twins, frequency is at chance while counterfactual attribution achieves ROC-AUC 0.744. Under a mean-preserving shape intervention, Wasserstein-1 and Jensen-Shannon evidence achieve ROC-AUC 0.909 and 0.967, while frequency and mean-based attribution remain at chance. Aggregate evidence complements repeated co-activity, improving mixed-mechanism ROC-AUC from 0.750 to 0.874 with a simple untrained combination.
发表机构
- University of Delaware（特拉华大学）
- Rochester Institute of Technology（罗切斯特理工学院）
机构由 AI 辅助整理，请以论文原文为准。
