跳转至

文章背景与核心概要

随着在线教育和远程考核的普及,传统的在线评估系统主要依赖浏览器锁定、网络摄像头监控以及行为分析等手段。然而,这些方法对于通过截图、屏幕共享、光学字符识别(OCR)以及自动化爬虫等手段直接窃取评估内容的攻击依然显得力不从心。为了应对这一挑战,本文引入了多层上下文伪装理论(Multi-Layer Context Camouflaging Theory, MCCT),作为多模态评估韧性套件(MARS)中多维时空上下文伪装模型(MSCCM)的重要扩展。

MCCT 的核心技术在于利用语义叠加,将真实的评估内容与合成的伪装内容融合成一个统一的渲染输出。该渲染内容对于人类合法用户而言依然清晰易读,但对于未经授权的自动化提取通道则会产生极高的计算模糊性。文章通过显式的提取通道算子对对抗性提取过程进行建模,构建了包含上下文反演算子、上下文叠层算子等六大耦合构件,并利用条件熵将计算模糊性公式化。该研究为行为自适应、兼顾无障碍访问且具备计算韧性的数字评估奠定了严谨的数学基础。


摘要 (Abstract)

Contemporary online assessment systems rely primarily on browser lockdown, webcam monitoring, and behavioural analytics, yet remain vulnerable to attacks that extract the assessment content itself through screenshots, screen sharing, optical character recognition, and automated scraping.

当代在线评估系统主要依赖浏览器锁定、网络摄像头监控和行为分析,但对于通过截图、屏幕共享、光学字符识别和自动化爬虫直接提取评估内容的攻击,依然存在漏洞。

This paper extends the Multi-dimensional Spatio-Temporal Context Camouflaging Model (MSCCM) within the MARS (Multi-modal Assessment Resilience Suite) by introducing the Multi-Layer Context Camouflaging Theory (MCCT), a mathematical framework that protects rendered assessment content through semantic superposition. Authentic assessment content and synthetically generated camouflage are represented as a unified rendering while remaining recoverable only by legitimate candidates.

本文扩展了多模态评估韧性套件(MARS)中的多维时空上下文伪装模型(MSCCM),引入了多层上下文伪装理论(MCCT)。这是一个通过语义叠加保护渲染评估内容的数学框架。真实的评估内容和合成生成的伪装内容被呈现为一个统一的渲染结果,同时仅能由合法候选人进行恢复。

The framework models the adversarial extraction process through an explicit extraction-channel operator and develops six coupled constructs: 1. The Context Inversion Operator 2. Contextual Lamination Operator 3. Separation Channel 4. Human Readability Functional 5. Computational Ambiguity Functional 6. Context Camouflage Tensor

该框架通过显式的提取通道算子对对抗性提取过程进行建模,并开发了六个耦合构件: 1. 上下文反演算子 2. 上下文叠层算子 3. 分离通道 4. 人类可读性泛函 5. 计算模糊性泛函 6. 上下文伪装张量

Computational ambiguity is formulated using conditional entropy, yielding a closed-form expression that quantifying uncertainty during unauthorized extraction, while legitimate recovery is guaranteed through an exact filtering identity. We further establish theoretical properties governing ambiguity, camouflage density, semantic preservation, multi-observation leakage, and temporal multiplexing, and present a rendering algorithm with computational complexity and a pre-registered evaluation protocol.

计算模糊性是通过条件熵建立公式的,从而导出一个闭式表达式,用于量化未经授权提取过程中的不确定性,同时通过精确的滤波恒等式保证合法恢复。我们进一步确立了控制模糊性、伪装密度、语义保持、多观测泄漏和时间复用的理论特性,并提出了具有计算复杂度和预注册评估协议的渲染算法。

MCCT provides a mathematically rigorous foundation for behaviorally adaptive, accessibility-aware, and computationally resilient digital assessment by securing rendered assessment content while preserving readability for legitimate users.

通过在保护渲染评估内容的同时保持合法用户的可读性,MCCT 为具备行为自适应、无障碍感知和计算韧性的数字评估提供了一个数学严谨的基础。


摘要与附加资源 (Metadata & Additional Resources)

  • arXiv Identifier: arXiv:2608.13100 [cs.AI]
  • Submitted on: August 13, 2026
  • Subjects: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Human-Computer Interaction (cs.HC)
  • Authors: Gupta Lovi Raj, Kaur Kamalpreet, Dama Sri Ram, Parani Prajithaa
  • Paper Metrics: 11 Pages, 36 Equations, 8 Figures
  • arXiv 标识符: arXiv:2608.13100 [cs.AI]
  • 提交时间: 2026年8月13日
  • 研究主题: 人工智能 (cs.AI);计算机与社会 (cs.CY);人机交互 (cs.HC)
  • 作者: Gupta Lovi Raj, Kaur Kamalpreet, Dama Sri Ram, Parani Prajithaa
  • 论文指标: 11页,36个方程式,8张图表

评论与全文访问 (Comments & Full-Text Access)

外部工具与参考资料 (External Tools & References)

外部工具与参考资料