评估法学硕士产出中的文化偏见和历史误解的框架

Moon-Kuen Mak , Tiejian Luo
{"title":"评估法学硕士产出中的文化偏见和历史误解的框架","authors":"Moon-Kuen Mak ,&nbsp;Tiejian Luo","doi":"10.1016/j.tbench.2025.100235","DOIUrl":null,"url":null,"abstract":"<div><div>Large Language Models (LLMs), while powerful, often perpetuate cultural biases and historical inaccuracies from their training data, marginalizing underrepresented perspectives. To address these issues, we introduce a structured framework to systematically evaluate and quantify these deficiencies. Our methodology combines culturally sensitive prompting with two novel metrics: the Cultural Bias Score (CBS) and the Historical Misconception Score (HMS). Our analysis reveals varying cultural biases across LLMs, with certain Western-centric models, such as Gemini, exhibiting higher bias. In contrast, other models, including ChatGPT and Poe, demonstrate more balanced cultural narratives. We also find that historical misconceptions are most prevalent for less-documented events, underscoring the critical need for training data diversification. Our framework suggests the potential effectiveness of bias-mitigation techniques, including dataset augmentation and human-in-the-loop (HITL) verification. Empirical validation of these strategies remains an important direction for future work. This work provides a replicable and scalable methodology for developers and researchers to help ensure the responsible and equitable deployment of LLMs in critical domains such as education and content moderation.</div></div>","PeriodicalId":100155,"journal":{"name":"BenchCouncil Transactions on Benchmarks, Standards and Evaluations","volume":"5 3","pages":"Article 100235"},"PeriodicalIF":0.0000,"publicationDate":"2025-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A framework for evaluating cultural bias and historical misconceptions in LLMs outputs\",\"authors\":\"Moon-Kuen Mak ,&nbsp;Tiejian Luo\",\"doi\":\"10.1016/j.tbench.2025.100235\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>Large Language Models (LLMs), while powerful, often perpetuate cultural biases and historical inaccuracies from their training data, marginalizing underrepresented perspectives. To address these issues, we introduce a structured framework to systematically evaluate and quantify these deficiencies. Our methodology combines culturally sensitive prompting with two novel metrics: the Cultural Bias Score (CBS) and the Historical Misconception Score (HMS). Our analysis reveals varying cultural biases across LLMs, with certain Western-centric models, such as Gemini, exhibiting higher bias. In contrast, other models, including ChatGPT and Poe, demonstrate more balanced cultural narratives. We also find that historical misconceptions are most prevalent for less-documented events, underscoring the critical need for training data diversification. Our framework suggests the potential effectiveness of bias-mitigation techniques, including dataset augmentation and human-in-the-loop (HITL) verification. Empirical validation of these strategies remains an important direction for future work. This work provides a replicable and scalable methodology for developers and researchers to help ensure the responsible and equitable deployment of LLMs in critical domains such as education and content moderation.</div></div>\",\"PeriodicalId\":100155,\"journal\":{\"name\":\"BenchCouncil Transactions on Benchmarks, Standards and Evaluations\",\"volume\":\"5 3\",\"pages\":\"Article 100235\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2025-08-18\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"BenchCouncil Transactions on Benchmarks, Standards and Evaluations\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S2772485925000481\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"BenchCouncil Transactions on Benchmarks, Standards and Evaluations","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2772485925000481","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

摘要

大型语言模型(llm)虽然强大,但往往会使其训练数据中的文化偏见和历史不准确性永久化,使未被充分代表的观点边缘化。为了解决这些问题,我们引入了一个结构化的框架来系统地评估和量化这些缺陷。我们的方法结合了文化敏感提示和两个新指标:文化偏见评分(CBS)和历史误解评分(HMS)。我们的分析揭示了不同法学硕士的文化偏见,某些以西方为中心的模式,如双子座,表现出更高的偏见。相比之下,其他模式,包括ChatGPT和Poe,展示了更平衡的文化叙事。我们还发现,对于记录较少的事件,历史误解最为普遍,这强调了训练数据多样化的关键需求。我们的框架表明了偏见缓解技术的潜在有效性,包括数据集增强和人在环路(HITL)验证。这些策略的实证验证仍然是未来工作的重要方向。这项工作为开发人员和研究人员提供了一种可复制和可扩展的方法,以帮助确保法学硕士在教育和内容审核等关键领域的负责任和公平部署。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
A framework for evaluating cultural bias and historical misconceptions in LLMs outputs
Large Language Models (LLMs), while powerful, often perpetuate cultural biases and historical inaccuracies from their training data, marginalizing underrepresented perspectives. To address these issues, we introduce a structured framework to systematically evaluate and quantify these deficiencies. Our methodology combines culturally sensitive prompting with two novel metrics: the Cultural Bias Score (CBS) and the Historical Misconception Score (HMS). Our analysis reveals varying cultural biases across LLMs, with certain Western-centric models, such as Gemini, exhibiting higher bias. In contrast, other models, including ChatGPT and Poe, demonstrate more balanced cultural narratives. We also find that historical misconceptions are most prevalent for less-documented events, underscoring the critical need for training data diversification. Our framework suggests the potential effectiveness of bias-mitigation techniques, including dataset augmentation and human-in-the-loop (HITL) verification. Empirical validation of these strategies remains an important direction for future work. This work provides a replicable and scalable methodology for developers and researchers to help ensure the responsible and equitable deployment of LLMs in critical domains such as education and content moderation.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
CiteScore
4.80
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信