From National Curricula to Cultural Awareness: Constructing Open-Ended Culture-Specific Question Answering Dataset

2026-01-08

arXiv.org

AI Summary

This study addresses the limited depth of cultural understanding exhibited by large language models (LLMs) in non-English, multicultural contexts. To bridge this gap, the authors propose CuCu, a novel multi-agent LLM collaboration framework that leverages national social studies curricula as supervised signals for cultural alignment. Using the Korean social studies curriculum as a foundation, they automatically generate a culturally grounded open-ended question-answering dataset, KCaQA, comprising 34.1k high-quality QA pairs covering uniquely Korean cultural themes. Empirical evaluation demonstrates that fine-tuning LLMs on KCaQA significantly enhances their awareness of and alignment with local sociocultural contexts, thereby advancing the capacity of language models to reason about culture-specific knowledge in a nuanced and contextually appropriate manner.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Large language models (LLMs) achieve strong performance on many tasks, but their progress remains uneven across languages and cultures, often reflecting values latent in English-centric training data. To enable practical cultural alignment, we propose a scalable approach that leverages national social studies curricula as a foundation for culture-aware supervision. We introduce CuCu, an automated multi-agent LLM framework that transforms national textbook curricula into open-ended, culture-specific question-answer pairs. Applying CuCu to the Korean national social studies curriculum, we construct KCaQA, comprising 34.1k open-ended QA pairs. Our quantitative and qualitative analyses suggest that KCaQA covers culture-specific topics and produces responses grounded in local sociocultural contexts.

Problem

• cultural alignment
• large language models
• culture-specific QA
• multilingual bias
• sociocultural context

Innovation

• culture-aware LLMs
• national curriculum
• multi-agent framework
• open-ended QA dataset
• cultural alignment