arXiv 2503.107892025-07-25cs.CL

数据漫画：论非裔美国人语言在预训练语料库中的表征

Data Caricatures: On the Representation of African American Language in Pretraining Corpora

Columbia University, Department of Computer Science（哥伦比亚大学计算机科学系）；University of Michigan, Department of Linguistics（密歇根大学语言学系）；University of Pennsylvania, School of Social Policy and Practice, Annenberg School for Communications（宾夕法尼亚大学社会政策与实践学院、安纳伯格通讯学院）；University of Tennessee, Knoxville, Department of English（田纳西大学基洛纳分校英语系）

Nicholas Deas, Blake Vente, Amith Ananthram, Jessica A. Grieser, Desmond Patton, Shana Kleiner, James Shepard, Kathleen McKeown

备注：: ACL 2025

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

该研究结合定量实验、人工判断与定性分析，考察12个开源英语预训练语料库中AAL的数量和质量，发现其严重不足、部分文本有害且过滤器偏向WME。

摘要：

通过结合定量实验、人类判断和定性分析，我们评估了12个以英语为主的开源预训练语料库中非裔美国人语言（AAL）表征的数量与质量。我们特别关注所收录的、代表AAL使用群体的AAL文本的来源、变异和自然度。我们发现，与美国人口统计数据相比，AAL在所有受评估的预训练语料库中均表征不足，仅占文档的0.007%至0.18%。我们还发现，C4中超过25%的AAL文本可能被认为不适合由LLM生成，并会强化有害刻板印象。最后，我们发现，大多数自动化过滤器在预训练语料库中更倾向于保留白人主流英语（WME）文本，而非AAL文本。

英文摘要：

With a combination of quantitative experiments, human judgments, and qualitative analyses, we evaluate the quantity and quality of African American Language (AAL) representation in 12 predominantly English, open-source pretraining corpora. We specifically focus on the sources, variation, and naturalness of included AAL texts representing the AAL-speaking community. We find that AAL is underrepresented in all evaluated pretraining corpora compared to US demographics, constituting as few as 0.007% and at most 0.18% of documents. We also find that more than 25% of AAL texts in C4 may be perceived as inappropriate for LLMs to generate and to reinforce harmful stereotypes. Finally, we find that most automated filters are more likely to conserve White Mainstream English (WME) texts over AAL in pretraining corpora.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。