Elamur Research⌕
☰
home / papers / 2026 / LLMs as Cultural Archives: Cultural Commonsense Knowledge Graph Extraction

LLMs as Cultural Archives: Cultural Commonsense Knowledge Graph Extraction

Paper · 2026 · Junior Cedric Tonga, Chen Cecilia Liu, Iryna Gurevych et al.

Summary

The paper asks whether the cultural knowledge that large language models absorb during pretraining can be turned into an explicit, structured resource. Existing cultural knowledge bases mostly store isolated facts and are usually in English. The authors instead build a Cultural Commonsense Knowledge Graph (CCKG) in the ATOMIC style, using iterative prompting of GPT-4o and Llama-3.3-70B-Instruct.

The method works in two stages:

  • Initial generation: for each subtopic, country and language, the model produces if–then assertions linking actions.
  • Iterative expansion: sequential links are re-prompted to add intermediate and follow-up steps, yielding multi-step chains of cultural practice.

Human annotators rated the graphs for five countries on cultural relevance, correctness and path coherence. Native-language extractions carried richer cultural detail, but English versions were more coherent and generally preferred, even for cultures that are not English-speaking. This points to uneven cultural encoding in current LLMs.

Supplying CCKG paths to smaller models improved cultural commonsense reasoning and story generation, with English chains giving the largest benefit. The work suggests that chain-structured knowledge is a usable basis for culturally grounded NLP.

Key points

  • Graph edges use five relation types: xNext, xEffect, xNeed, oNext and oEffect. Assertions are composed into paths of actions, such as an Indonesian breakfast sequence involving lontong, soto ayam and a warung.

  • Downstream reasoning was evaluated on the IndoCulture and ArabCulture benchmarks. Story generation used Llama3.1-8B-IT, Qwen2.5-7B-IT and Gemma2-9B-IT across three countries.

  • In English story generation, adding CCKG paths improved quality for all nine model–country pairs. Llama models gained about +1.4 in Cultural Relevance on average, while Fluency and Coherence improved less.

  • GPT-4o used as a judge agreed with human raters much better on native-language stories than on English ones:

    MetricEnglish storiesNative-language stories
    Cultural relevance~0.4~0.8
    Fluency−0.10.9
    Coherence0.00.8

  • Code is released on GitHub (JuniorTonga/Cultural_Commonsense_Knowledge_Graph). The authors present the extracted content as a research prototype rather than a formal dataset and caution about possible stereotypes.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Technische Universität DarmstadtDarmstadt · Germany109 works · 35,920 citations🔬Hessian Center for Artificial IntelligenceDarmstadt · Germany12 works · 529 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

🧭 Topics 3Research subjects and languages the work is about.Language models & pretraining · 3,320 worksReasoning · 1,426 worksKnowledge graphs & ontologies · 192 works

Cite

Cite as Junior Cedric Tonga, Chen Cecilia Liu, Iryna Gurevych and Fajri Koto. 2026. LLMs as Cultural Archives: Cultural Commonsense Knowledge Graph Extraction. Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.eacl-long.295

@inproceedings{tonga2026llms,
  title = {{LLMs as Cultural Archives: Cultural Commonsense Knowledge Graph Extraction}},
  author = {Junior Cedric Tonga, Chen Cecilia Liu, Iryna Gurevych and Fajri Koto},
  booktitle = {Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.eacl-long.295},
  url = {https://doi.org/10.18653/v1/2026.eacl-long.295}
}

References

45 references, 22 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Muhammad Farid Adilazuarda, Sagnik Mukherjee, Pradhyumna Lavania, Siddhant Shivdutt Singh, Alham Fikri Aji, Jacki O’Neill, Ashutosh Modi, and Monojit Choudhury. 2024. Towards measuring and modeling “culture” in LLMs: A survey. In Proceedings of the 2024 Conference on Empirical Methods in Natural Lan…
  2. Mubashara Akhtar, Abhilash Shankarampeta, Vivek Gupta, Arpit Patil, Oana Cocarascu, and Elena Simperl. 2023. Exploring the numerical reasoning capabilities of language models: A comprehensive analysis on tabular data. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 15…
  3. Alec Radford, Ilya Sutskever, and Dario Amodei. 2020. Language models are few-shot learners. In Advances in Neural Information Processing Systems 33 (NeurIPS 2020), pages 1877–1901. Curran Associates, Inc.
  4. Courtney Corley and Rada Mihalcea. 2005. Measuring the semantic similarity of texts. In Proceedings of the ACL Workshop on Empirical Modeling of Semantic Equivalence and Entailment, pages 13–18, Ann Arbor, Michigan. Association for Computational Linguistics.
  5. Li Du, Xiao Ding, Kai Xiong, Ting Liu, and Bing Qin. 2022. e-CARE: a new dataset for exploring explainable causal reasoning. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 432–446, Dublin, Ireland. Association for Computation… find
  6. Saeed Almheiri, Rania Elbadry, Mena Attia, Chenxi Wang, Preslav Nakov, Timothy Baldwin, and Fajri Koto. 2025. Cross-cultural transfer of commonsense reasoning in LLMs: Evidence from the Arab world. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 4593–4614, Suzhou, Chi… find
  7. Esin Durmus, Karina Nguyen, Thomas Liao, Nicholas Schiefer, Amanda Askell, Anton Bakhtin, Carol Chen, Zac Hatfield-Dodds, Danny Hernandez, Nicholas Joseph, et al. 2024. Towards measuring the representation of subjective global opinions in language models. In First Conference on Language Modeling.
  8. Mehar Bhatia and Vered Shwartz. 2023. GD-COMET: A geo-diverse commonsense inference model. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 7993–8001, Singapore. Association for Computational Linguistics. find
  9. Yi Fung, Ruining Zhao, Jae Doo, Chenkai Sun, and Heng Ji. 2024. Massively multi-cultural knowledge acquisition & LM benchmarking. arXiv preprint, abs/2402.09369.
  10. Yonatan Bisk, Rowan Zellers, Ronan Le Bras, Jianfeng Gao, and Yejin Choi. 2020. PIQA: reasoning about physical commonsense in natural language. In The Thirty-Fourth AAAI Conference on Artificial Intelligence, AAAI 2020, The Thirty-Second Innovative Applications of Artificial Intelligence Conference,…
  11. Antoine Bosselut, Hannah Rashkin, Maarten Sap, Chaitanya Malaviya, Asli Celikyilmaz, and Yejin Choi. 2019. COMET: Commonsense transformers for automatic knowledge graph construction. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 4762–4779, Florence…
  12. Tom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D. Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel M. Ziegler, Jeffrey Wu, Clemens Winte… — The Llama 3 Herd of Models
  13. Christian Haerpfer, Ronald Inglehart, Alejandro Moreno, Christian Welzel, Kseniya Kizilova, Jaime Diez-Medrano, Marta Lagos, Pippa Norris, Eduard Ponarin, Bjorn Puranen, et al. 2022. World values survey: Round seven–country-pooled datafile version 5.0. find
  14. Daniel Hershcovich, Stella Frank, Heather Lent, Miryam de Lhoneux, Mostafa Abdou, Stephanie Brandl, Emanuele Bugliarello, Laura Cabello Piqueras, Ilias Chalkidis, Ruixiang Cui, Constanza Fierro, Katerina Margatina, Phillip Rust, and Anders Søgaard. 2022. Challenges and strategies in cross-cultural N…
  15. Jing Huang and Diyi Yang. 2023. Culturally aware natural language inference. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 7591–7609, Singapore. Association for Computational Linguistics.
  16. Jena D. Hwang, Chandra Bhagavatula, Ronan Le Bras, Jeff Da, Keisuke Sakaguchi, Antoine Bosselut, and Yejin Choi. 2021. (comet-) atomic 2020: On symbolic and neural commonsense knowledge graphs. In Thirty-Fifth AAAI Conference on Artificial Intelligence, AAAI 2021, Thirty-Third Conference on Innovati…
  17. Zhengbao Jiang, Jun Araki, Haibo Ding, and Graham Neubig. 2021. How can we know when language models know? on the calibration of language models for question answering. Transactions of the Association for Computational Linguistics, 9:962–977. find
  18. Fajri Koto, Rahmad Mahendra, Nurul Aisyah, and Timothy Baldwin. 2024. IndoCulture: Exploring geographically influenced cultural commonsense reasoning across eleven Indonesian provinces. Transactions of the Association for Computational Linguistics, 12:1703–1719.
  19. Haitao Li, Qian Dong, Junjie Chen, Huixue Su, Yujia Zhou, Qingyao Ai, Ziyi Ye, and Yiqun Liu. 2024. Llms-as-judges: A comprehensive survey on llm-based evaluation methods. arXiv preprint, abs/2412.05579.
  20. Bill Yuchen Lin, Seyeon Lee, Rahul Khanna, and Xiang Ren. 2020. Birds have four legs?! NumerSense: Probing Numerical Commonsense Knowledge of Pre-Trained Language Models. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 6862–6868, Online. Assoc…
  21. Chen Cecilia Liu, Iryna Gurevych, and Anna Korhonen. 2025. Culturally aware and adapted NLP: A taxonomy and a survey of the state of the art. Transactions of the Association for Computational Linguistics, 13:652–689.
  22. Tarek Naous, Michael J Ryan, Alan Ritter, and Wei Xu. 2024. Having beer after prayer? measuring cultural bias in large language models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 16366–16393, Bangkok, Thailand. Associatio…
  23. Tuan-Phong Nguyen, Simon Razniewski, Aparna S. Varde, and Gerhard Weikum. 2023. Extracting cultural commonsense knowledge at scale. In Proceedings of the ACM Web Conference 2023, WWW 2023, Austin, TX, USA, 30 April 2023 - 4 May 2023, pages 1907–1917. ACM. find
  24. Tuan-Phong Nguyen, Simon Razniewski, and Gerhard Weikum. 2024. Cultural commonsense knowledge for intercultural dialogues. In Proceedings of the 33rd ACM International Conference on Information and Knowledge Management, CIKM ’24, page 1774–1784, New York, NY, USA. Association for Computing Machinery…
  25. OpenAI, :, Aaron Hurst, Adam Lerer, Adam P. Goucher, Adam Perelman, Aditya Ramesh, Aidan Clark, AJ Ostrow, Akila Welihinda, Alan Hayes, Alec Radford, and Aleksander M adry et al. 2024. Gpt-4o system card. arXiv preprint, abs/2410.21276.
  26. Haoyi Qiu, Alexander Fabbri, Divyansh Agarwal, Kung-Hsiang Huang, Sarah Tan, Nanyun Peng, and Chien-Sheng Wu. 2025. Evaluating cultural and social awareness of LLM web agents. In Findings of the Association for Computational Linguistics: NAACL 2025, pages 3978–4005, Albuquerque, New Mexico. Associat… find
  27. Nils Reimers and Iryna Gurevych. 2019. Sentence-BERT: Sentence embeddings using Siamese BERT-networks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 3982–3992,…
  28. Julien Romero, Simon Razniewski, Koninika Pal, Jeff Z. Pan, Archit Sakhadeo, and Gerhard Weikum. 2019. Commonsense properties from query logs and question answering forums. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management, CIKM ’19, page 1411–1420, New… find
  29. Abdelrahman Sadallah, Junior Cedric Tonga, Khalid Almubarak, Saeed Almheiri, Farah Atif, Chatrine Qwaider, Karima Kadaoui, Sara Shatnawi, Yaser Alesh, and Fajri Koto. 2025. Commonsense reasoning in Arab culture. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguisti… find
  30. Maarten Sap, Ronan Le Bras, Emily Allaway, Chandra Bhagavatula, Nicholas Lourie, Hannah Rashkin, Brendan Roof, Noah A. Smith, and Yejin Choi. 2019a. ATOMIC: an atlas of machine commonsense for if-then reasoning. In The Thirty-Third AAAI Conference on Artificial Intelligence, AAAI 2019, The Thirty-Fi… find
  31. Maarten Sap, Hannah Rashkin, Derek Chen, Ronan Le Bras, and Yejin Choi. 2019b. Social IQa: Commonsense reasoning about social interactions. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Process… find
  32. Maarten Sap, Vered Shwartz, Antoine Bosselut, Yejin Choi, and Dan Roth. 2020. Commonsense reasoning for natural language processing. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics: Tutorial Abstracts, pages 27–33, Online. Association for Computational Ling… find
  33. An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, Huan Lin, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jiaxi Yang, Jingren Zhou, Junyang Lin, Kai Dang, Keming Lu, Keqin Bao, Kexin Yang, Le Yu, Mei Li, Mingfeng Xue, Pei… find
  34. Weiyan Shi, Ryan Li, Yutong Zhang, Caleb Ziems, Sunny Yu, Raya Horesh, Rogério Abreu De Paula, and Diyi Yang. 2024. CultureBank: An online community-driven knowledge base towards culturally aware language technologies. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 4…
  35. Da Yin, Hritik Bansal, Masoud Monajatipoor, Liunian Harold Li, and Kai-Wei Chang. 2022. GeoM-LAMA: Geo-diverse commonsense probing on multilingual pre-trained language models. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 2039–2055, Abu Dhabi, Unite… find
  36. Robyn Speer, Joshua Chin, and Catherine Havasi. 2017. Conceptnet 5.5: An open multilingual graph of general knowledge. In Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence, February 4-9, 2017, San Francisco, California, USA, pages 4444–4451. AAAI Press. find
  37. Tianyi Zhang*, Varsha Kishore*, Felix Wu*, Kilian Q. Weinberger, and Yoav Artzi. 2020. Bertscore: Evaluating text generation with bert. In 8th International Conference on Learning Representations, ICLR 2020, Addis Ababa, Ethiopia, April 26-30, 2020. OpenReview.net. find
  38. Kai Sun, Yifan Xu, Hanwen Zha, Yue Liu, and Xin Luna Dong. 2024. Head-to-tail: How knowledgeable are large language models (LLMs)? A.K.A. will LLMs replace knowledge graphs? In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human La… find
  39. A We recruited two annotators per country (ten in total) and applied strict eligibility requirements to ensure cultural authenticity and linguistic proficiency. Annotators were required to meet the following criteria: • Native speaker of the specified local language and proficient in English (speaki… find
  40. Qingyu Tan, Hwee Tou Ng, and Lidong Bing. 2023. Towards benchmarking and improving the temporal reasoning capability of large language models. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 14820–14835, Toronto, Canada. Assoc… find
  41. Resided in the country for at least ten years. • Demonstrated deep familiarity with the country’s culture. find
  42. Niket Tandon, Gerard de Melo, Fabian M. Suchanek, and Gerhard Weikum. 2014. Webchild: harvesting and organizing commonsense knowledge from the web. Proceedings of the 7th ACM international conference on Web search and data mining. find
  43. Gemma Team, Morgane Riviere, Shreya Pathak, Pier Giuseppe Sessa, Cassidy Hardin, Surya Bhupatiraju, Léonard Hussenot, and Thomas Mesnard et al. 2024. Gemma 2: Improving open language models at a practical size. arXiv preprint, abs/2408.00118. find
  44. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Ichter, Fei Xia, Ed H. Chi, Quoc V. Le, and Denny Zhou. 2022. Chain-of-thought prompting elicits reasoning in large language models. In Proceedings of the 36th International Conference on Neural Information Processing Systems, NIPS ’22, R…
  45. Annotator Criteria • Both parents are also natives residing in the same country. • Minimum qualification of senior high school graduation (higher degrees preferred). find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/tonga-2026-llms-as-cultural-archives-cultural/