Jokes Aside: Measuring the Semantic Distance of Double Meanings
作者: Fabio De Ponte
分类: cs.CL
发布日期: 2026-08-21
备注: The paper was submitted to ISHS (International Society for Humor Studies) conference held in Kraków, Poland on 7-11 July 2025. It was awarded the GSA AWARD and was presented during a special plenary session (see the section Graduate Student Awards, 2006-2025 of the webpage https://www.humorstudies.org/ConferCenter.htm)
💡 一句话要点
提出基于语义距离的新指标以改进幽默生成模型
🎯 匹配领域: 支柱六：视频提取与匹配 (Video Extraction) 支柱九：具身大模型 (Embodied Foundation Models)
关键词: 幽默生成 语义距离 上下文嵌入 对称性度量 机器学习 自然语言处理 双关语
📋 核心要点
- 现有的幽默生成模型在预测幽默评分时表现不佳，尤其是在语义距离的度量上存在不足。
- 本文提出了新的对称性度量，并重新审视了基于上下文嵌入向量的三个已有度量，以改进幽默生成模型的效果。
- 实验结果显示，尽管新模型的预测准确率低于基线，但对称性度量与高评分笑话之间存在显著关联，提示其在幽默生成中的潜在价值。
📝 摘要（中文）
大型语言模型显著丰富了计算幽默研究的工具，尤其是在自动生成笑话和双关语方面。本文基于Petrovic和Matthews（2013）提出的笑话生成模型，重新审视了三个度量标准，并引入了对称性度量。研究使用OpenAI和MiniLM模型在三个数据集上收集嵌入向量，结果显示，基于新度量的模型在幽默评分预测上表现不佳，但对称性度量与高评分笑话存在一致关联，表明其可能是幽默的必要属性。
🔬 方法详解
问题定义：本文旨在解决现有幽默生成模型在预测幽默评分时的不足，特别是对语义距离的度量不够准确的问题。
核心思路：论文通过引入对称性度量，结合上下文嵌入向量，重新审视已有的幽默度量标准，以期提高幽默生成模型的表现。
技术框架：研究使用OpenAI和MiniLM模型在JokeJudger、Expunations和rJokes三个数据集上收集嵌入向量，分析不同度量对幽默评分的影响。
关键创新：引入对称性度量作为新指标，强调Z与X和Y的接近性，这一创新与现有方法的本质区别在于更好地捕捉幽默的语义关系。
关键设计：在模型训练中，使用了OpenAI text-embedding-3-small和MiniLM all-MiniLM-L6-v2作为嵌入向量生成器，并扩展了数据集以包含双重含义的句子对。
🖼️ 关键图片
📊 实验亮点
实验结果表明，基于新度量的模型在JokeJudger数据集上的最佳准确率为57.1%，低于61.5%的基线，而在Expunations和rJokes上的表现更低。尽管如此，对称性度量与高评分笑话之间的关联性提示其在幽默生成中的重要性。
🎯 应用场景
该研究的潜在应用领域包括自动化幽默生成、社交媒体内容创作和人机交互系统。通过改进幽默生成模型，可以提升用户体验，增加互动的趣味性，未来可能在娱乐和教育领域产生深远影响。
📄 摘要（原文）
Large language models have significantly enriched the toolkit for computational humor research, particularly in the automated generation of jokes and puns. A key innovation, contextual embedding vectors, offers new opportunities to revisit and refine earlier hypotheses. Notably, Petrovic and Matthews (2013) proposed a joke generation model based on the scheme "I like my X like I like my Y, Z" (e.g. "I like my ice like I like my dreams, crushed"). They suggested that joke hilarity increases with: a) frequent association of Z with X and Y, b) rarity of Z, c) ambiguity of Z, and d) meaning distance between X and Y. Building on this, Winters et al. (2019) proposed a set of metrics, based on Google Ngrams and Word2Vector. In this work, three out of their five metrics are revisited with word embeddings: obviousness, compatibility, and comparison. Another measure, symmetry, defined as closeness of Z to both X and Y, is introduced here for the first time. Two models were used to collect the embedding vectors (OpenAI text-embedding-3-small and MiniLM all-MiniLM-L6-v2) on three datasets: JokeJudger, Expunations, and rJokes. The last two datasets, Expunations, and rJokes, were expanded by adding paired sentences that captured the ambiguous expression at the core of each joke in its two different meanings. Results revealed that models trained on the proposed metrics performed poorly in predicting humor ratings: on JokeJudger, the best model achieved 57.1% accuracy, below the 61.5% baseline, while performance on Expunations and rJokes was even lower. Nevertheless, the symmetry metric seems consistently associated with higher-rated jokes, suggesting it may capture a necessary -though not sufficient- property of humor.