用少量标注数据高效检测外交对话中的谎言，关键在识别稀有欺骗信号。
PU-Lie: Lightweight Deception Detection in Imbalanced Diplomatic Dialogues via Positive-Unlabeled Learning
- 基于冻结BERT+语言与游戏特征，结合正负样本学习框架。
- 在仅5%标签数据下实现0.60的宏平均F1，参数量减少650倍。
- 适合资源有限且需精准识别欺骗的高风险场景应用。
在战略对话中检测欺骗是一项复杂且高风险的任务，因语言微妙且欺骗与真实信息之间存在极端类别不平衡。本文重新审视Diplomacy数据集中的欺骗检测问题，其中少于5%的消息被标注为欺骗。我们提出一种轻量但有效的模型PU-Lie，结合冻结的BERT嵌入、可解释的语言与游戏特定特征以及正-未标记（PU）学习目标。不同于传统二分类器，PU-Lie专为仅有少量欺骗消息标注而多数未标注的情况设计。模型在7种模型的综合评估与消融实验中，实现了0.60的新最优宏平均F1，同时将可训练参数减少超过650倍。研究证明了PU学习、语言可解释性及说话人感知表示的价值。特别强调，在该设定下，准确识别欺骗比识别真实消息更为关键，这一优先级驱动了对PU学习的选择，其显式建模稀有但重要的欺骗类别。
原文摘要 · Abstract (English)
Detecting deception in strategic dialogues is a complex and high-stakes task due to the subtlety of language and extreme class imbalance between deceptive and truthful communications. In this work, we revisit deception detection in the Diplomacy dataset, where less than 5% of messages are labeled deceptive. We introduce a lightweight yet effective model combining frozen BERT embeddings, interpretable linguistic and game-specific features, and a Positive-Unlabeled (PU) learning objective. Unlike traditional binary classifiers, PU-Lie is tailored for situations where only a small portion of deceptive messages are labeled, and the majority are unlabeled. Our model achieves a new best macro F1 of 0.60 while reducing trainable parameters by over 650x. Through comprehensive evaluations and ablation studies across seven models, we demonstrate the value of PU learning, linguistic interpretability, and speaker-aware representations. Notably, we emphasize that in this problem setting, accurately detecting deception is more critical than identifying truthful messages. This priority guides our choice of PU learning, which explicitly models the rare but vital deceptive class.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。