Detecting Deceptive Recruitment: A Signal-theoretic Machine Learning Framework for Early Identification of Labour Exploitation
作者: Sajid Siraj, Mahnaz Hosseinzadeh, Amin Vafadarnikjoo, Shuyang Li
分类: cs.CY, cs.LG
发布日期: 2026-09-17
💡 一句话要点
提出信号理论机器学习框架以检测欺诈性招聘问题
🎯 匹配领域: 支柱九:具身大模型 (Embodied Foundation Models)
关键词: 欺诈性招聘 信号理论 多模态检测 计算机视觉 自然语言处理 风险评估 机器学习
📋 核心要点
- 现有方法在检测欺诈性招聘广告方面面临数据稀缺和缺乏有效指标的挑战,导致系统性检测手段不足。
- 论文提出将欺诈性招聘检测视为信号理论下的分类问题,利用多模态数据结合计算机视觉和自然语言处理进行分析。
- 实验结果表明,单一模态的区分能力显著,ROC-AUC值在0.87到0.97之间,集成模型虽提升有限,但仍具实用价值。
📝 摘要(中文)
欺诈性在线招聘广告已成为强迫劳动的主要途径,但由于数据稀缺和缺乏经验证的指标,系统检测方法仍然不够成熟。本文将这一检测挑战形式化为信号理论下的分类问题,利用464个经过验证的案例(164个欺诈性,300个合法)开发了多模态检测模型,结合计算机视觉、自然语言处理和语义嵌入。通过系统的特征消融实验和重复的分层交叉验证,证明了各个模态具有显著的区分能力(ROC-AUC: 0.87--0.97),而它们的集成则带来了适度的进一步提升。SHAP分析显示,文本质量和领域特定的风险语言是主要区分因素,生产质量差距反映了资源限制,阻止了剥削者在所有沟通渠道上同时保持专业标准。
🔬 方法详解
问题定义:本文旨在解决欺诈性招聘广告的检测问题,现有方法因数据稀缺和缺乏有效指标而难以系统化。
核心思路:将欺诈性招聘检测形式化为信号理论下的分类问题,利用多模态数据(文本、视觉、结构)进行综合分析,以识别剥削者的虚假信号。
技术框架:整体架构包括数据收集、特征提取、模型训练和评估四个主要阶段。数据来自反奴隶制慈善机构,涵盖多个国家和行业。
关键创新:通过结合计算机视觉和自然语言处理,提出了一种新的多模态检测模型,显著提高了欺诈性招聘广告的识别能力,与传统方法相比具有更高的准确性和鲁棒性。
关键设计:模型采用SHAP分析确定关键特征,文本质量、风险语言密度和可读性指数被认为是主要区分因素,视觉特征如颜色和纹理也被纳入考虑。
🖼️ 关键图片
📊 实验亮点
实验结果显示,单一模态的ROC-AUC值在0.87到0.97之间,表明各模态具有显著的区分能力。尽管集成模型的提升有限,但仍然为实际应用提供了有效的决策支持。
🎯 应用场景
该研究的潜在应用领域包括人力资源管理、在线招聘平台和反剥削组织。通过提供可解释的风险评分,帮助从业者识别和防范欺诈性招聘广告,具有重要的社会价值和实际影响。
📄 摘要(原文)
Deceptive online job advertisements have emerged as a primary pathway into forced labour, yet systematic detection methods remain underdeveloped due to data scarcity and absence of empirically validated indicators. We formalise this detection challenge as a classification problem under signalling theory, where exploiters transmit costless signals mimicking legitimate communications across textual, visual, and structural dimensions. Using 464 verified cases (164 deceptive, 300 legitimate) collected through anti-slavery charities across nine origin countries and 21 industries, we develop multimodal detection models combining computer vision, natural language processing, and semantic embeddings. Through systematic feature ablation experiments and repeated stratified cross-validation, we demonstrate that individual modalities achieve substantial discriminatory power (ROC-AUC: 0.87--0.97), whilst their integration yields modest further gains. SHAP-based analysis reveals that text quality and domain-specific risk language are the primary discriminators, with readability indices, risk keyword density, and visa sponsorship mentions ranking highest, followed by visual colour and texture features. These production quality gaps reflect resource constraints that prevent exploiters from maintaining professional standards across all communication channels simultaneously. We operationalise findings through a proof-of-concept decision support system providing interpretable risk scores for practitioners. This work demonstrates how rigorous analytical frameworks can address complex humanitarian operations challenges characterised by information asymmetry and limited ground-truth data.