Elamur Research⌕
☰
home / papers / 2026 / Studying Expert-ese: Profiling and Classification of Domain-Specific Language Variation in Architecture with Traditional Machine Learning and LLMs

Studying Expert-ese: Profiling and Classification of Domain-Specific Language Variation in Architecture with Traditional Machine Learning and LLMs

Summary

The paper asks whether professional training in architecture leaves detectable marks on how people speak spontaneously. The authors treat this as "expert-ese", a domain register comparable to translationese. They use the ExpLay Corpus, which contains German oral descriptions of architectural images by trained architects and by laypeople. Profiling-UD and the DECAF framework are used to extract syntactic, morpho-syntactic and raw-text features. These features yield a linguistic profile for each group and a corpus-level divergence score. The authors then train several classifiers to separate expert from non-expert speech: two feature-based models, a small BiLSTM, and two LLMs. The groups differ only slightly overall. Even so, the BiLSTM with fastText embeddings separates them best, which suggests that semantic cues carry more signal than structural ones and that compact neural models handle small datasets well. The results support the Grammar of Space hypothesis. They also point to differences in how experts package information, beyond jargon alone, which matters for expert–layperson communication in fields such as architecture, medicine or law.

Key points

  • Data: ExpLay Corpus, which has matched elicitation conditions and is released under CC BY 4.0 via OSF.
  • Corpus divergence: pairwise Jensen–Shannon divergence between the expert and lay subcorpora is 0.25. PCA and ANOVA explained little variance, and no single variable was significant.
  • Expert profile: experts used longer tokens, longer sentences and greater lexical diversity, which is consistent with denser content-word use.
  • Classification: logistic regression and SVM were compared with a BiLSTM and the LLMs GliClass and LLaMA 2. The fastText-based BiLSTM scored best, with F1 = 0.88.
  • Limitations: the dataset is small, which also rules out fine-tuning. Automatic parsing may introduce errors, and only one domain was studied. Future work includes human classification tests and information-density analyses.

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

Fields: Linguistics & languages · History & humanities

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.latechclfl-1.3

@inproceedings{schacht2026studying,
  title = {{Studying Expert-ese: Profiling and Classification of Domain-Specific Language Variation in Architecture with Traditional Machine Learning and LLMs}},
  author = {Schacht, Carmen and Danhier, Renate Delucchi},
  booktitle = {Proceedings of the 10th Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature 2026},
  year = {2026},
  doi = {10.18653/v1/2026.latechclfl-1.3},
  url = {https://doi.org/10.18653/v1/2026.latechclfl-1.3}
}

References

57 references, 5 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Shlomo Engelson Argamon. 2019. Computational register analysis and synthesis. ArXiv, abs/1901.02543. — [arXiv:1901.02543](https://arxiv.org/abs/1901.02543) find
  2. Douglas Biber and Susan Conrad. 2009. Register, Genre, and Style. Cambridge Textbooks in Linguistics. Cambridge University Press. find
  3. Douglas Biber and Bethany Gray. 2016. Grammatical Complexity in Academic English: Linguistic Change in Writing. Studies in English Language. Cambridge University Press. find
  4. Douglas Biber, Bethany Gray, and Kornwipa Poonpon. 2011. Should we use characteristics of conversation to measure grammatical complexity in l2 writing development? TESOL Quarterly, 45(1):5–35. \#\# Limitations While this study provides insights into linguistic variation between experts and non-experts… find
  5. Par-Dominique Brunato, Andrea Cimino, Felice Dell’Orletta, Giulia Venturi, and Simonetta Montemagni. 2020. Profiling-UD: a tool for linguistic profiling of texts. In Proceedings of the Twelfth Language Resources and Evaluation Conference, pages 7145–7151, Marseille, France. European Language Resourc… find
  6. Andrea Cimino, Martijn Wieling, Felice Dell’Orletta, Simonetta Montemagni, and Giulia Venturi. 2017. Identifying predictive features for textual genre classification: the key role of syntactic features. In Proceedings of the Fourth Italian Conference on Computational Linguistics (CLiC-it 2017 ), CEU… find
  7. John C. B. Gamboa, Leigh B. Fernandez, and Shanley E. M. Allen. 2024. Investigating the uniform information density hypothesis with complex nominal compounds. Applied Psycholinguistics, 45(2):322–367. find
  8. Kevyn Collins-Thompson. 2014. Computational assessment of text readability: A survey of current and future research. ITL - International Journal of Applied Linguistics, 165(2):97–135. find
  9. Edouard Grave, Piotr Bojanowski, Prakhar Gupta, Armand Joulin, and Tomas Mikolov. 2018. Learning word vectors for 157 languages. In Proceedings of the International Conference on Language Resources and Evaluation (LREC 2018). find our card
  10. M. Gellerstam. 1986. Translationese in swedish novels translated from english. In L. Wollin and H. Lindquist, editors, Translation studies in Scandinavia, pages 88–95. CWK Gleerup, Lund. find
  11. Marie-Catherine de Marneffe, Christopher D. Manning, Joakim Nivre, and Daniel Zeman. 2021. Universal Dependencies. Computational Linguistics, 47(2):255–308. find our card
  12. M. A. K. Halliday. 1988/2004. On the language of physical science. In Jonathan J. Webster, editor, The Collected Works of M. A. K. Halliday (Vol. 5), pages 140–158. Continuum, London and New York. find
  13. Stefania Degaetano-Ortlieb. 2021. Measuring informativity: The rise of compounds as informationally dense structures in 20th century Scientific English, pages 291–312. John Benjamins Publishing Company. find
  14. Pengcheng He, Xiaodong Liu, Jianfeng Gao, and Weizhu Chen. 2020. Deberta: Decoding-enhanced bert with disentangled attention. Https://api.semanticscholar.org/CorpusID:219531210. find our card
  15. Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long short-term memory. Neural Computation, 9(8):1735– 1780. find
  16. Stefania Degaetano-Ortlieb and Andrew Piper. 2019. The scientization of literary study. In Proceedings of the 3rd Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature, pages 18–28, Minneapolis, USA. Association for Computational Linguis… find
  17. T. F. Jaeger. 2010. Redundancy and reduction: Speakers manage syntactic information density. Cognitive Psychology, 61(1):23–62. find
  18. Deborah N. Jakobi, Thomas Kern, David R. Reich, Patrick Haller, and Lena A. Jäger. 2024. Potec: A german naturalistic eye-tracking-while-reading corpus. Preprint, arXiv:2403.00506. find
  19. Stefania Degaetano-Ortlieb and Elke Teich. 2018. Using relative entropy for detection and analysis of periods of diachronic linguistic change. In Proceedings of the Second Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature, pages 22–3… find
  20. Chunzhen Jiang, Aritra Kundu, and Mark Claypool. 2020. Game player response times versus task dexterity and decision complexity. In Extended Abstracts of the 2020 Annual Symposium on Computer-Human Interaction in Play, CHI PLAY ’20, page 277–281, New York, NY, USA. Association for Computing Machiner… find
  21. Stefania Degaetano-Ortlieb and Elke Teich. 2022. Toward an optimal code for communication: The case of scientific english. Corpus Ling.. Ling.. Theory, 18(1):175–207. find
  22. I. T. Jolliffe. 2002. Principal Component Analysis. Springer Series in Statistics. Springer-Verlag, New York. Doi:10.1007/b98835. — [doi:10.1007/b98835](https://doi.org/10.1007/b98835) find
  23. Renate Delucchi Danhier, Barbara Mertins, Holger Mertins, and Gerold Schneider. 2025. Entropy as a lens: Exploring visual behavior patterns in architects. Journal of Eye Movement Research, 18(5). find
  24. Tom S Juzek, Marie-Pauline Krielke, and Elke Teich. 2020. Exploring diachronic syntactic shifts with dependency length: the case of scientific English. In Proceedings of the Fourth Workshop on Universal Dependencies (UDW 2020), pages 109–119, Barcelona, Spain (Online). Association for Computational… find our card
  25. A. Ersin, H. Ceren Tezeren, N. Ozunlu Pekyavas, B. Asal, A. Atabey, A. Diri, and İ Gonen. 2022. The relationship between reaction time and gaming time in e-sports players. Kinesiology, 54(1):36–42. Doi:10.26582/k.54.1.4. — [doi:10.26582/k.54.1.4](https://doi.org/10.26582/k.54.1.4) find
  26. S Kullback and R A Leibler. 1951. On information and sufficiency. The Annals of Mathematical Statistics, 22(1):79–86. find
  27. John Gamboa, Kristina Braun, Juhani Järvikivi, and Shanley E. M. Allen. 2025. The distributional properties of long nominal compounds in scientific articles: an investigation based on the uniform information density hypothesis. Corpus Linguistics and Linguistic Theory, 21(1):137–171. find
  28. M. Kunilovskaya and G. Corpas Pastor. 2021. Translationese and register variation in english-to-russian professional translation. In V. X. Wang, L. Lim, and D. Li, editors, New Perspectives on Corpus Translation Studies. New Frontiers in Translation Studies. , . https. Springer, Singapore. Doi:10.10… — [doi:10.1007/978-981-16-4918-9_6](https://doi.org/10.1007/978-981-16-4918-9_6) find
  29. F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot, and E. Duchesnay. 2011. Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12:282… find
  30. Kristopher Kyle. 2016. Measuring Syntactic Development in L2 Writing: Fine Grained Indices of Syntactic Complexity and Usage-Based Indices of Syntactic Sophistication. Ph.D. thesis, Georgia State University. Doi:10.57709/8501051. — [doi:10.57709/8501051](https://doi.org/10.57709/8501051) find
  31. Jeffrey Pennington, Richard Socher, and Christopher D. Manning. 2014. Glove: Global vectors for word representation. In Empirical Methods in Natural Language Processing (EMNLP), pages 1532–1543. find
  32. Willem J. M. Levelt. 1989. Speaking: From Intention to Articulation. The MIT Press. find
  33. S. C. Levinson. 2003. Space in language and cognition: Explorations in cognitive diversity. Cambridge University Press. find
  34. Simon Perkins, Kevin Lacker, and James Theiler. 2003. Grafting: Fast, incremental feature selection by gradient descent in function space. The Journal of Machine Learning Researchs, 3:1333–1356. find
  35. Xiaofei Lu. 2010. Automatic analysis of syntactic complexity in second language writing. International Journal of Corpus Linguistics, 15(4):474–496. find
  36. D. Roberson, I. Davies, and J. Davidoff. 2000. Color categories are not universal: Replications and new evidence from a stone-age culture. Journal of Experimental Psychology: General, 129(3):369–398. find
  37. E. A. Maguire, D. G. Gadian, I. S. Johnsrude, C. D. Good, J. Ashburner, R. S. J. Frackowiak, and C. D. Frith. 2000. Navigation-related structural change in the hippocampi of taxi drivers. Proceedings of the National Academy of Sciences, 97(8):4398–4403. find
  38. Raphael Rubino, Stefania Degaetano-Ortlieb, Elke Teich, and Josef van Genabith. 2016. Modeling diachronic change in scientific writing with information density. In Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers, pages 750–761, Osaka, Japa… find
  39. The COLING 2016 Organizing Committee. find
  40. A. Majid, M. Bowerman, S. Kita, D. B. Haun, and S. C. Levinson. 2004. Can language restructure cognition? the case for space. Trends in Cognitive Sciences, 8(3):108–114. find
  41. Carmen Schacht and Renate Delucchi Danhier. 2025. ExpLay: A new corpus resource for the research on expertise as an influential factor on language production. In Proceedings of the 19th Linguistic Annotation Workshop (LAW-XIX-2025), pages 216–227, Vienna, Austria. Association for Computational Lingu… find
  42. H. Mertins, R. Delucchi Danhier, B. Mertins, A. Schulz, and B. Schulz. 2020. The role of expertise in the perception of architectural space. In C. Leopold, C. Robeller, and U. (Hrsg. Weber, editors, Research Culture in Architecture, pages 279–288. Birkhäuser, Basel. find
  43. Claude E. Shannon. 1948. A mathematical theory of communication. The Bell System Technical Journal, 27(3):379–423. find
  44. Kanishka Misra and Kyle Mahowald. 2024. Language models learn rare phenomena from less rare phenomena: The case of the missing AANNs. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 913–929, Miami, Florida, USA. Association for Computational Linguisti… find
  45. D. I. Slobin. 1996. From “thought and language” to “thinking for speaking". In J. J. Gumperz and S. C. Levinson, editors, Rethinking linguistic relativity, pages 70–96. Cambridge University Press. find
  46. Ihor Stepanov, Mykhailo Shtopko, Dmytro Vodianyt-skyi, Oleksandr Lukashov, Alexander Yavorskyi, and Mykyta Yaroshenko. 2025. Gliclass: Generalist lightweight model for sequence classification tasks. Preprint, arXiv:2508.07662. — [arXiv:2508.07662](https://arxiv.org/abs/2508.07662) find
  47. Max Müller-Eberstein, Rob Van Der Goot, and Anna Rogers. 2025. DECAF: A dynamically extensible corpus analysis framework. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations), pages 351–362, Vienna, Austria. Association for Comp… find
  48. A. Papafragou, J. Hulbert, and J. Trueswell. 2008. Does language guide event perception? evidence from eye movements. Cognition, 108(1):155–184. find
  49. Elke Teich, Stefania Degaetano-Ortlieb, Peter Fankhauser, Hannah Kermes, and Ekaterina Lapshinova-Koltunski. 2016. The linguistic construal of disciplinarity: A data-mining approach using register features. Journal of the Association for Information Science and Technology, 67(7):1668–1678. find
  50. Shinjae Park. 2024. Identifying key linguistic variables of second language speaking proficiency using principal component analysis. Forum for Linguistic Studies, 6(6):623–633. find
  51. Elke Teich, Peter Fankhauser, Stefania Degaetano-Ortlieb, and Yuri Bizzoni. 2021. Less is more/more diverse: On the communicative utility of linguistic conventionalization. Frontiers in Communication, 5. find
  52. Polarity, Poss, PronType, Reflex, Style, Tense, Typo, VerbForm, Voice, deprel, upos, and xpos. Out of the full set, the top six contributing features are displayed in Table 2. Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Pra… find
  53. Wes McKinney. 2010. Data Structures for Statistical Computing in Python. In Proceedings of the 9th Python in Science Conference, pages 56 – 61. find
  54. Benjamin Lee Whorf. 1956. Language, Thought, and Reality. Cambridge, Ma. find
  55. J. Winawer, N. Witthoft, M. C. Frank, L. Wu, A. R. Wade, and L. Boroditsky. 2007. Russian blues reveal effects of language on color discrimination. Proceedings of the National Academy of Sciences, 104(19):7780–7785. find
  56. Contribution score Person xpos PronType VerbForm deprel upos 0.070 0.050 0.050 0.050 0.050 0.040 The contribution of the use of person in the two sub-corpora is pointed out in Table 3. *Table 3: Contribution of the use of person in both groups.* A K Wong and M You. 1985. Entropy and distance of rand… find
  57. Group First Person Third Person 0.135 0.071 0.864 0.928 Experts Non-experts Menglin Xia, Ekaterina Kochmar, and Ted Briscoe. 2016. Text readability assessment for second language learners. In Proceedings of the 11th Workshop on Innovative Use of NLP for Building Educational Applications, pages 12–22… find our card

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/schacht-2026-studying-expert-ese-profiling-classification/