Elamur Research⌕
☰
home / papers / 2026 / Ecological Discourse Modeling in a Low-Resource Setting: A Longitudinal Vietnamese Climate Corpus with Comparative Topic Modeling

Ecological Discourse Modeling in a Low-Resource Setting: A Longitudinal Vietnamese Climate Corpus with Comparative Topic Modeling

Paper · 2026 · Huyen Phuong Nguyen

Summary

Most computational work on climate discourse covers English and other high-resource languages. This paper builds a longitudinal Vietnamese climate news corpus and uses it to study how national media coverage is organised thematically and how it has changed over two decades. The author scraped 10,401 articles (April 2004 – February 2026) from VTV, VnExpress and Nhân Dân using keyword queries. Text was segmented with VnCoreNLP, and transformer-based NER was then applied to keep only 4,501 documents focused on Vietnam. This subset was analysed with LDA and BERTopic, and the topics were scored for coherence (c_v, NPMI) and diversity. Both models recover consistent themes, including pollution, renewable energy and climate governance. Over time, coverage moves from incident-driven pollution stories toward policy, adaptation and energy-transition framing. BERTopic yields more coherent topics at similar diversity. The author also argues that co-occurrence-based coherence scores may undervalue topic quality in Vietnamese. The work provides a reusable resource and methodological guidance for analysing ecological discourse in low-resource languages.

Key points

  • Corpus and release: 10,401 articles from three mainstream outlets, spanning 23 years. Only URLs, titles and dates are released, at github.com/hnnphuongg/vietnam-climate-news, because copyright prevents sharing the full text.
  • Preprocessing: VnCoreNLP handles multi-syllable word segmentation (e.g., khí_hậu). Terms found in fewer than 10 documents or in over 70% of documents are dropped. NER filtering produces the 4,501-document Vietnam-focused subset.
  • LDA themes: the yearly mean document–topic probability was used to follow six topics, namely Air Quality, Climate Adaptation, Environmental Pollution, Energy Transition, Policy & International, and Renewable Projects.
  • Temporal patterns:
    • Pollution coverage spikes around 2006–07, in 2016 (the Formosa Ha Tinh incident) and in 2019–20 (urban air quality).
    • Adaptation coverage peaks near 2012.
    • Energy coverage follows a U-shape, rising from 2016 onward.
  • Limitations: the corpus leans toward mainstream national media, segmentation may introduce noise, and topics were not validated by human judges.

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

🎓Toulouse School of EconomicsToulouse · France2 works · 1 citation

 Fields, topics and concepts

📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 worksSocial Sciences · 1,954 worksGeneral Social Sciences · 254 works

🧭 Topics 2Research subjects and languages the work is about.Discourse analysis · 471 worksLow-resource languages · 287 works

💡 Concepts 1Key ideas and methods explained on our concept pages.Topic modeling · 47 works

Cite

Cite as Huyen Phuong Nguyen. 2026. Ecological Discourse Modeling in a Low-Resource Setting: A Longitudinal Vietnamese Climate Corpus with Comparative Topic Modeling. Proceedings of the 2nd Workshop on Ecology, Environment, and Natural Language Processing.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.63317/4n8q3zryqvju

@inproceedings{nguyen2026ecological,
  title = {{Ecological Discourse Modeling in a Low-Resource Setting: A Longitudinal Vietnamese Climate Corpus with Comparative Topic Modeling}},
  author = {Huyen Phuong Nguyen},
  booktitle = {Proceedings of the 2nd Workshop on Ecology, Environment, and Natural Language Processing},
  year = {2026},
  doi = {10.63317/4n8q3zryqvju},
  url = {https://doi.org/10.63317/4n8q3zryqvju}
}

References

20 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Leland McInnes and John Healy. 2018. Umap: Uniform manifold approximation and projection for dimension reduction.
  2. Rafiu Adekoya Badekale and Adewale Akinfaderin. 2025. Temporal analysis of climate policy discourse: Insights from dynamic embedded topic modeling. ArXiv, abs/2507.06435. — [arXiv:2507.06435](https://arxiv.org/abs/2507.06435) find
  3. Leland McInnes, John Healy, and Steve Astels. 2017. hdbscan: Hierarchical density based clustering. The Journal of Open Source Software, 2:205.
  4. David M. Blei, Andrew Y. Ng, and Michael I. Jordan. 2003. Latent dirichlet allocation. Journal of Machine Learning Research, 3:993–1022.
  5. Dasha Pruss, Yoshinari Fujinuma, Ashlynn R. Daughton, Michael J. Paul, Brad Arnot, Danielle Albers Szafir, and Jordan Boyd-Graber. 2019. Zika discourse in the americas: A multilingual topic analysis of twitter. PloS one, 14(5):e0216922. find
  6. Gerlof Bouma. 2009. Normalized (pointwise) mutual information in collocation extraction. Proceedings of the Biennial GSCL Conference 2009. find
  7. Michael Röder, Andreas Both, and Alexander Hinneburg. 2015. Exploring the space of topic coherence measures. WSDM ’15, page 399–408, New York, NY, USA. Association for Computing Machinery.
  8. Ziqiang Cao, Sujian Li, Yang Liu, Wenjie Li, and Heng Ji. 2015. A novel neural topic model and its supervised extension. In Proceedings of the Twenty-Ninth AAAI Conference on Artificial Intelligence (AAAI-15), pages 2210–2216. AAAI Press. find
  9. Sahil Sawant, Jinhong Yu, Kirtikumar Pandya, Chun-Kit Ngan, and Rolf Bardeli. 2022. An enhanced bertopic framework and algorithm for improving topic coherence and diversity. In Proceedings of the 2022 IEEE 24th International Conference on High Performance Computing & Communications (HPCC/DSS/SmartCi… find
  10. Rob Churchill and Lisa Singh. 2022. The evolution of topic modeling. ACM Computing Surveys, 54(10s):1–35. find
  11. Thi Kim Phung Dang. 2025. Climate change communication in vietnam’s online newspapers and its implications for climate actions. Sustainability, 17:1354. find
  12. Ehsan Tashakori, Yaser Sobhanifard, Adel Aazami, and Rahim Khanizad. 2025. Uncovering semantic patterns in sustainability research: A systematic nlp review. Sustainable Development. find
  13. Adji B. Dieng, Francisco J. R. Ruiz, and David M. Blei. 2020. Topic modeling in embedding spaces. Transactions of the Association for Computational Linguistics, 8:439–453.
  14. Samson Ebenezar Uthirapathy and Domnic Sandanam. 2023. Topic modelling and opinion analysis on climate change twitter data using lda and bert model. Procedia Computer Science, 218:908–917. find
  15. Isabella Gagliardi and Teresa Artese. 2020. Semantic unsupervised automatic keyphrase extraction by integrating word embedding with clustering methods. Multimodal Technologies and Interaction, 4(2):30. find
  16. Azmine Toushik Wasi, Wahid Faisal, Taj Ahmad, Abdur Rahman, and Mst Rafia Islam. 2024. Dhoroni: Exploring bengali climate change and environmental views with a multi-perspective news dataset and natural language processing. arXiv preprint arXiv:2410.17225. — [arXiv:2410.17225](https://arxiv.org/abs/2410.17225) find
  17. Tunahan Gokcimen and Bihter Das. 2024. Exploring climate change discourse on social media and blogs using a topic modeling analysis. Heliyon, 10(11):e32464. find
  18. Maarten R. Grootendorst. 2022. Bertopic: Neural topic modeling with a class-based tf-idf procedure. ArXiv, abs/2203.05794.
  19. F. M. Anamul Haque, Md. Abdur Rahman, and Sumon Ahmed. 2025. Ghtm: A graph based hybrid topic modeling approach in low-resource bengali language. ArXiv, abs/2508.00605. find
  20. Long Le and Lien-Huong Vo. 2026. Vietnam in the climate change narratives: A discursive news values analysis of english-language news. Studies in Media and Communication, 14:267–280. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/nguyen-2026-ecological-discourse-modeling-low-resource/