Elamur Research⌕
☰
home / papers / 2026 / Benchmarking NLP-supported Language Sample Analysis for Swiss Children's Speech

Benchmarking NLP-supported Language Sample Analysis for Swiss Children's Speech

Paper · 2026 · Anja Ryser, Yingqiang Gao and Sarah Ebling

Summary

Language sample analysis (LSA) helps speech-language pathologists diagnose developmental language disorder (DLD) in children. It is rarely used in practice because the manual work is heavy. The authors test whether locally deployed NLP tools, with no commercial LLMs, can make LSA semi-automatic for children in German-speaking Switzerland. They recorded therapist–child interactions with 119 children aged four to eight, with typical and atypical development, in Swiss German dialects and in Swiss Standard German. Trained students produced reference transcripts in an adapted Dieth spelling, and the authors also created orthographically normalized versions. Against these references, they evaluated zero-shot Whisper transcription and POS tagging with UPOS and STTS tags. They also outline a pipeline and software in which the clinician corrects the transcript and then receives measures such as mean length of utterance and subject–verb agreement. Performance is uneven: error rates are much higher for Swiss German and for children's speech, and normalization improves both ASR and tagging. The study shows that privacy-friendly NLP support for LSA is feasible. Clinical-grade quality will still require dedicated training and fine-tuning.

Key points

  • Data collection: 25 speech-language pathologists held 10–20-minute sessions with 119 children. Swiss German accounts for 91 recordings (about 18 h, 126,733 words) and Swiss Standard German for 19 recordings (about 3.5 h, 21,356 words). Consent to publish was obtained for 41 recordings.
  • ASR setup: The model was Whisper (openai/whisper-small), used zero-shot. On original transcripts, average WER was 81.0 for Swiss German and 48.7 for Swiss Standard German. Normalization lowered these to 57.2 and 45.6.
  • Speaker groups: ASR errors and POS tagging quality were consistently better for the adult therapists than for the children. For example, UPOS F1 on normalized Swiss German was 0.835 for adults and 0.787 for children.
  • Pipeline and software: The proposed workflow keeps the human expert in the loop. It runs from recording to automatic transcription, then manual correction, then automatic feature profiling (POS tags, morphology, dependencies, SVA). The pathologist makes the final diagnostic decision.
  • Released resource: A small case-study dataset of six annotated transcripts in Swiss German and Swiss Standard German is to be published on SWISSUbase (DOI 10.48656/rqf2-sq76).

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

🎓University of ZurichZurich · Switzerland78 works · 12,260 citations

 Fields, topics and concepts

📚 Shelves 8The 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 worksPsychology · 1,552 worksDevelopmental and Educational Psychology · 499 worksMedicine · 241 worksPhysiology · 29 works

💡 Concepts 1Key ideas and methods explained on our concept pages.Latent semantic indexing · 29 works

Cite

Cite as Anja Ryser, Yingqiang Gao and Sarah Ebling. 2026. Benchmarking NLP-supported Language Sample Analysis for Swiss Children's Speech. Proceedings of the Sixth Resources and ProcessIng of linguistic, para-linguistic and extra-linguistic Data from people with various forms of cognitive/psychiatric/developmental impairments in cooperation with the MENTAL.ai consortium.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.63317/3wbvp3v9kj68

@inproceedings{ryser2026benchmarking,
  title = {{Benchmarking NLP-supported Language Sample Analysis for Swiss Children's Speech}},
  author = {Anja Ryser, Yingqiang Gao and Sarah Ebling},
  booktitle = {Proceedings of the Sixth Resources and ProcessIng of linguistic, para-linguistic and extra-linguistic Data from people with various forms of cognitive/psychiatric/developmental impairments in cooperation with the MENTAL.ai consortium},
  year = {2026},
  doi = {10.63317/3wbvp3v9kj68},
  url = {https://doi.org/10.63317/3wbvp3v9kj68}
}

References

53 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Noëmi Aepli and Rico Sennrich. 2022. Improving Zero-shot Cross-lingual Transfer Between Closely Related Languages by Injecting Character-level Noise. In Findings of the Association for Computational Linguistics: ACL 2022, pages 4074–4083. find
  2. Junyi Ao, Rui Wang, Long Zhou, Chengyi Wang, Shuo Ren, Yu Wu, Shujie Liu, Tom Ko, Qing Li, Yu Zhang, et al. 2022. SpeechT5: Unified-modal Encoder-decoder Pre-training for Spoken Language Processing. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1:…
  3. Pelin Dogan-Schönberger, Julian Mäder, and Thomas Hofmann. 2021. Swissdial: Parallel Multidialectal Corpus of Spoken Swiss German. arXiv preprint arXiv:2103.11401. find
  4. Eyal Liron Dolev, Clemens Fidel Lutz, and Noëmi Aepli. 2024. Does Whisper Understand Swiss German? An Automatic, Qualitative, and Human Evaluation. arXiv preprint arXiv:2404.19310. find
  5. Yuriy Arabskyy, Aashish Agarwal, Subhadeep Dey, and Oscar Koller. 2021. Dialectal Speech Recognition and Translation of Swiss German Speech to Standard German Text: Microsoft’s Submission to SwissText 2021. arXiv preprint arXiv:2106.08126. find
  6. Keyur Gabani, Thamar Solorio, Yang Liu, Khairunnisa Hassanali, and Christine A Dollaghan. 2011. Exploring A Corpus-based Approach for Detecting Language Impairment in Monolingual English-speaking Children. Artificial Intelligence in Medicine, 53(3):161–170. find
  7. Rebecca Bawayan, Jennifer A Brown, Rebecca Bawayan, and Jennifer A Brown. 2022. Language Sample Analysis Consideration and Use: A Survey of School-based Speech Language Pathologists. Clinical Archives of Communication Disorders, 7(1):15–28. find
  8. John F Gallagher and Jill R Hoover. 2020. Measure What You Treat: Using Language Sample Analysis for Grammatical Outcome Measures in Children with Developmental Language Disorder. Perspectives of the ASHA Special Interest Groups, 5(2):350–363. find
  9. Vivek Bhardwaj, Mohamed Tahar Ben Othman, Vinay Kukreja, Youcef Belkhier, Mohit Bajaj, B Srikanth Goud, Ateeq Ur Rehman, Muhammad Shafiq, and Habib Hamam. 2022. Automatic Speech Recognition (ASR) Systems for Children: A Systematic Literature Review. Applied Sciences, 12(9):4419. find
  10. Khairun-nisa Hassanali, Yang Liu, and Thamar Solorio. 2012. Evaluating NLP Features for Automatic Prediction of Language Impairment Using Child Speech Transcripts. In INTERSPEECH, pages 1339–1342. find
  11. Nora Hollenstein and Noëmi Aepli. 2014. Compilation of a Swiss German Dialect Corpus and Its Application to POS Tagging. In Proceedings of the first workshop on applying NLP tools to similar languages, varieties and dialects, pages 85–94. find
  12. Dorothy VM Bishop. 2006. What Causes Specific Language Impairment in Children? Current directions in psychological science, 15(5):217– 221. find
  13. Bernd Bohnet, Ryan McDonald, Gonçalo Simões, Daniel Andor, Emily Pitler, and Joshua Maynez. 2018. Morphosyntactic Tagging with a Meta-BiLSTM Model over Context Sensitive Token Encodings. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers… find
  14. Rishabh Jain, Andrei Barcovschi, Mariam Yiwere, Peter Corcoran, and Horia Cucu. 2023. Adaptation of Whisper Models to Child Speech Recognition. arXiv preprint arXiv:2307.13008. — [arXiv:2307.13008](https://arxiv.org/abs/2307.13008) find
  15. Tannon Kew, Iuliia Nigmatulina, Lorenz Nagele, and Tanja Samardzic. 2020. UZH TILT: A Kaldi Recipe for Swiss German Speech to Standard German Text. In SwissText/KONVENS. find
  16. Thorsten Brants. 2000. Inter-annotator Agreement for a German Newspaper Corpus. In Proceedings of the Second International Conference on Language Resources and Evaluation (LREC’00). find
  17. Inge S Klatte, Vera Van Heugten, Rob Zwitserlood, and Ellen Gerrits. 2022. Language Sample Analysis in Clinical Practice: Speech-language Pathologists’ Barriers, Facilitators, and Needs. Language, Speech, and Hearing Services in Schools, 53(1):1–16. find
  18. Burcu Can. 2011. Statistical Models for Unsupervised Learning of Morphology and POS Tagging. Ph.D. thesis, University of York. find
  19. Matej Klemen, Luka Krsnik, and Marko Robnik-Šikonja. 2023. Enhancing Deep Neural Networks with Morphological Information. Natural Language Engineering, 29(2):360–385. find
  20. Jacob Cohen. 1968. Weighted Kappa: Nominal Scale Agreement Provision for Scaled Disagreement or Partial Credit. Psychological bulletin, 70(4):213. find
  21. Dancheng Liu, Jason Yang, Ishan Albrecht-Buehler, Helen Qin, Sophie Li, Yuting Hu, Amir Nassereldine, and Jinjun Xiong. 2024. find
  22. Automatic Eugen Dieth. 1986. Schwyzertütschi Dialäktschrift. Cornelsen Schweiz AG. find
  23. Screening for Children with Speech Disorder using Automatic Speech Recognition: Opportunities and Challenges. In Proceedings of the AAAI Symposium Series, volume 4, pages 308–313. Corpus. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 3250–3256. find
  24. Michel Plüss, Lukas Neukom, Christian Scheller, and Manfred Vogel. 2020. Swiss Parliaments Corpus, An Automatically Aligned Swiss German Speech to Standard German Text Corpus. arXiv preprint arXiv:2010.02810. find
  25. Ulrike Lüdtke, Juan Bornman, Febe De Wet, Ulrich Heid, Jörn Ostermann, Lars Rumberg, Jeannie Van der Linde, and Hanna Ehlert. 2023. Multidisciplinary Perspectives on Automatic Analysis of Children’s Language Samples: Where Do We Go From Here? Folia Phoniatrica et Logopaedica, 75(1):1–12. find
  26. Niclas Pokel, Pehuén Moure, Roman Boehringer, and Yingqiang Gao. 2025. Adapting Foundation Speech Recognition Models to Impaired Speech: A Semantic Re-chaining Approach for Personalization of German Speech. In 12th edition of the Disfluency in Spontaneous Speech Workshop (DiSS 2025), pages 82–86. find
  27. Carina Lüke, Christina Kauschke, Andrea Dohmen, Andrea Haid, Christina Leitinger, Claudia Männel, Tanja Penz, Steffi Sachse, Wiebke Scharff Rethfeldt, Julia Spranger, et al. 2023. Definition and Terminology of Developmental Language Disorders—interdisciplinary Consensus Across German-speaking Countr… find
  28. Alexandros Potamianos, Shrikanth S Narayanan, and Sungbok Lee. 1997. Automatic Speech Recognition for Children. In Eurospeech, volume 97, pages 2371–2374. find
  29. Jon Miller, Robin Chapman, et al. 1985. Systematic Analysis of Language Transcripts. Madison, WI: Language Analysis Laboratory. find
  30. Clifton Pye. 1994. The CHILDES Project: Tools for Analyzing Talk. find
  31. Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, and Ilya Sutskever. 2023. Robust Speech Recognition via Large-scale Weak Supervision. In International conference on machine learning, pages 28492–28518. PMLR.
  32. Iuliia Nigmatulina, Tannon Kew, and Tanja Samardzic. 2020. ASR for Non-standardised Languages with Dialectal Variation: the Case of Swiss German. In Proceedings of the 7th Workshop on NLP for Similar Languages, Varieties and Dialects, pages 15–24. find
  33. Robert E Owens Jr, Stacey L Pavelko, and Dennis Bambinelli. 2018. Moving Beyond Mean Length of Utterance: Analyzing Language Samples to Identify Intervention Targets. Perspectives of the ASHA Special Interest Groups, 3(1):5–22. find
  34. Michelle N Ramos, Penelope Collins, and Elizabeth D Peña. 2022. Sharpening Our Tools: A Systematic Review to Identify Diagnostically Accurate Language Sample Measures. Journal of Speech, Language, and Hearing Research, 65(10):3890–3907. find
  35. Shantipriya Parida, Esaú Villatoro-Tello, Sajit Kumar, Petr Motlicek, and Qingran Zhan. 2020. Idiap Submission to Swiss-German Language Detection Shared Task. In SwissText/KONVENS, page 7. find
  36. Michelle Nichols Ramos. 2024. Using Language Sample Analysis to Identify Developmental Language Disorder in Bilingual Children. Ph.D. thesis, University of California, Irvine. find
  37. Yi Ren, Chenxu Hu, Xu Tan, Tao Qin, Sheng Zhao, Zhou Zhao, and Tie-Yan Liu. 2021. Fast-Speech 2: Fast and High-quality End-to-end Text to Speech. In International Conference on Learning Representations. find
  38. Michel Plüss, Jan Milan Deriu, Yanick Schraner, Claudio Paonessa, Julia Hartmann, Larissa Schmidt, Christian Scheller, Manuela Huerlimann, Tanja Samardzic, Manfred Vogel, et al. 2023. STT4SG-350: A Speech Corpus for All Swiss German Dialect Regions. In Proceedings of the 61st Annual Meeting of the A… find
  39. Lars Rumberg, Hanna Ehlert, Ulrike Lüdtke, and Jörn Ostermann. 2021. Age-Invariant Training for End-to-end Child Speech Recognition Using Adversarial Multi-task Learning. In Interspeech, pages 3850–3854. find
  40. Michel Plüss, Manuela Huerlimann, Marc Cuny, Alla Stöckli, Nikolaos Kapotis, Julia Hartmann, Malgorzata Anna Ulasik, Christian Scheller, Yanick Schraner, Amit Jain, et al. 2022. DS-200: A Swiss German Speech to Standard German Text Alessandra Sansavini, Maria Elena Favilla, Maria Teresa Guasti, Andr… find
  41. Children. Journal of Speech, Language, and Hearing Research, 39(6):1284–1294. find
  42. Hawau Toyin, Hao Li, and Hanan Aldarmaki. 2024. STTATTS: Unified Speech-to-text And Text-to-speech Model. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 6853–6863. find
  43. Vivian van Wijngaarden, Hester de Wilde, Dieuwke Mink van der Molen, Jildo Petter, Inge Stegeman, Ellen Gerrits, Adriana L Smit, and Marie-José van den Boogaard. 2024. Genetic Outcomes in Children with Developmental Language Disorder: A Systematic Review. Frontiers in pediatrics, 12:1315229. find
  44. Yanick Schraner, Christian Scheller, Michel Plüss, and Manfred Vogel. 2022. Swiss German Speech to Text System Evaluation. arXiv preprint arXiv:2207.00412. find
  45. Miikka Silfverberg and Krister Lindén. 2011. Combining Statistical Models for POS Tagging using Finite-state Calculus. In Proceedings of the 18th Nordic Conference of Computational Linguistics (NODALIDA 2011), pages 183–190. find
  46. Daniel V Smith, Alex Sneddon, Lauren Ward, Andreas Duenser, Jill Freyne, David Silvera-Tawil, and Angela Morgan. 2017. Improving Child Speech Disorder Assessment by Incorporating Out-of-domain Adult Speech. In Interspeech, pages 2690–2694. find
  47. Thamar Solorio. 2013. Survey on Emerging Research on the Use of Natural Language Processing in Clinical Language Assessment of Children. Language and Linguistics Compass, 7(12):633– 646. find
  48. Samuel Stucki, Mark Cieliebak, and Jan Deriu. 2025. SwissGPC v1. 0–The Swiss German Podcasts Corpus. arXiv preprint arXiv:2509.19866. find
  49. Elina Stüssi and Phillip Ströbel. 2024. Part-of-speech Tagging of 16th-Century Latin with GPT. In Proceedings of the 8th Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature (LaTeCH-CLfL 2024), pages 196–206. find
  50. Vincenzo Timmel, Claudio Paonessa, Reza Kakooee, Manfred Vogel, and Daniel Perru-choud. 2024. Fine-tuning Whisper on Low-Resource Languages for Real-World Applications. arXiv preprint arXiv:2412.15726. find
  51. Alexander Tkachenko and Kairit Sirts. 2018. Modeling Composite Labels for Neural Morphological Tagging. In Proceedings of the 22nd Conference on Computational Natural Language Learning, pages 368–379. find
  52. J Bruce Tomblin, Nancy L Records, and Xuyang Zhang. 1996. A System for the Diagnosis of Specific Language Impairment in Kindergarten A. find
  53. Instructions for Human Annotation of Part-of-speech Tagging on Gold Transcriptions (translated from the original German file) find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/ryser-2026-benchmarking-nlp-supported-language-sample/