---
title: "Revisiting German Complex Word Identification: Contextualized"
description: "Earlier studies also suggested that LLMs perform only modestly in this setting. 29 references."
authors: ["Thorben Schomacker", "Seid Muhie Yimam", "Chris Biemann", "Marina Tropmann-Frick"]
year: 2026
venue: "Proceedings of the Joint Workshop on Readability and Text Simplification (READIxTSAR) @ LREC 2026"
source: "https://aclanthology.org/2026.readi-1.1/"
doi: "10.63317/4z33nthezmjs"
licence: "CC BY-NC 4.0 (ELRA, LREC proceedings)"
licence_url: "https://creativecommons.org/licenses/by-nc/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/schomacker-2026-revisiting-german-complex-word-identification/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/schomacker-2026-revisiting-german-complex-word-identification/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/47/) / Revisiting German Complex Word Identification: Contextualized LLMs and Feature Injection

# Revisiting German Complex Word Identification: Contextualized LLMs and Feature Injection

**Authors:** Thorben Schomacker, [Seid Muhie Yimam](https://md.research.elamur.ai/free/en/authors/seid-muhie-yimam/), [Chris Biemann](https://md.research.elamur.ai/free/en/authors/chris-biemann/) and Marina Tropmann-Frick · **Year:** 2026 · **Venue:** [Proceedings of the Joint Workshop on Readability and Text Simplification (READIxTSAR) @ LREC 2026](https://md.research.elamur.ai/free/en/venues/readi/) ([Joint Workshop on Readability and Text Simplification (READIxTSAR) @ LREC 2026](https://md.research.elamur.ai/free/en/venues/readi/2026/)) · **Publisher:** [European Language Resources Association (ELRA)](https://md.research.elamur.ai/free/en/publishers/european-language-resources-association-elra/) · **Licence:** [CC BY-NC 4.0 (ELRA, LREC proceedings)](https://creativecommons.org/licenses/by-nc/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.readi-1.1/) · [PDF](https://aclanthology.org/2026.readi-1.1.pdf) · [DOI 10.63317/4z33nthezmjs](https://doi.org/10.63317/4z33nthezmjs) · [OpenAlex](https://openalex.org/W7166551565)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
German complex word identification (CWI), the task of detecting the words or phrases that are likely to hinder comprehension and should be simplified first, has received far less research attention than English CWI. Earlier studies also suggested that LLMs perform only modestly in this setting. The authors revisit the problem using the German part of the CWIG3G2 dataset from the CWI 2018 shared task, which is the only openly available word-level German CWI resource. They fine-tune three current models, two LLaMA-style and one BERT-based, on the probabilistic labels. Three input variants are compared: the target alone, the target tagged inside its sentence, and the tagged sentence plus classical length and frequency features written into the tag. Binary predictions come from a decision threshold tuned on the development set.

Models without fine-tuning lag behind, but fine-tuned models surpass 14 earlier baselines on both binary and probabilistic CWI. Sentence context improves results, while the injected features make no clear difference. The work matters because one released checkpoint serves both evaluation views, which gives a reusable building block for German text simplification.

## Key points
- **Data:** The CWIG3G2 German portion contains 7905 instances, which may be single words or multiword expressions. It keeps the shared-task split of 6151 train, 795 dev and 959 test instances.
- **Labels:** Each target is rated by 12 annotators, drawn from a pool of 23 (12 native, 11 non-native speakers). A target counts as binary-complex if at least one annotator flagged it, which corresponds to a probabilistic score of at least 1/12 ≈ 0.083.
- **Injected features:**
  - vowel count (vow) and character length (len);
  - syllable count (syl), computed with spaCy's syllables module;
  - a frequency class from the DWDS API (dwds), which returns None for multiword expressions.
- **Research questions:**
  - whether sentence context helps (it does);
  - whether injected features help (no measurable gain);
  - how sensitive binary performance is to the choice of threshold.
- **Released resources:** Code is on GitHub (tschomacker/german-cwi-llm), and model checkpoints are in a Hugging Face collection under the same name.

*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

- [Universität Hamburg](https://md.research.elamur.ai/free/en/universities/universitat-hamburg/) — University, Hamburg, Germany · 81 works · 5,549 citations

Also on the byline: [HAW Hamburg](https://ror.org/00fkqwx76).

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 18,614 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 17,965 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 17,450 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 17,296 works

🧭 Topics (1): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,894 works

## Cite

Cite as: Thorben Schomacker, Seid Muhie Yimam, Chris Biemann and Marina Tropmann-Frick. 2026. Revisiting German Complex Word Identification: Contextualized LLMs and Feature Injection. Proceedings of the Joint Workshop on Readability and Text Simplification (READIxTSAR) @ LREC 2026.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{schomacker2026revisiting,
  title = {{Revisiting German Complex Word Identification: Contextualized LLMs and Feature Injection}},
  author = {Thorben Schomacker, Seid Muhie Yimam, Chris Biemann and Marina Tropmann-Frick},
  booktitle = {Proceedings of the Joint Workshop on Readability and Text Simplification (READIxTSAR) @ LREC 2026},
  year = {2026},
  doi = {10.63317/4z33nthezmjs},
  url = {https://doi.org/10.63317/4z33nthezmjs}
}
```

- [← previous Revisiting Faithfulness Annotations for Long-form Summaries](https://md.research.elamur.ai/free/en/papers/zhong-2026-revisiting-faithfulness-annotations-long-form/)
- [next → Revisiting Non-Verbatim Memorization in Large Language Models: The Role of Entity Surface Forms](https://md.research.elamur.ai/free/en/papers/nishida-2026-revisiting-non-verbatim-memorization-large/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/47/)

## References

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

1. Miriam Anschütz, Joshua Oehms, Thomas Wimmer, Bart\lomiej Jezierski, and Georg Groh. 2023. Language Models for German Text Simplification: Overcoming Parallel Data Scarcity through Style-specific Pre-training. In Findings of the Association for Computational Linguistics: ACL 2023, pages 1147–1158, Toronto, Canada. Association for Computational Linguistics. URL: https://aclanthology.org/2023.findings-acl.74, doi:10.18653/v1/2023. findings-acl.74. [↗](https://doi.org/10.18653/v1/2023)
2. Joachim Bingel and Johannes Bjerva. 2018. Cross-lingual complex word identification with multitask learning. In Proceedings of the Thirteenth Workshop on Innovative Use of NLP for Building Educational Applications, pages 166–174, New Orleans, Louisiana, USA. Association for Computational Linguistics. URL: https://aclanthology.org/W18-0518/, doi:10.18653/v1/W18-0518. [↗](https://doi.org/10.18653/v1/w18-0518)
3. Silvana Deilen, Sergio Hernández Garrido, Ekaterina Lapshinova-Koltunski, and Christiane Maaß. 2023. Using ChatGPT as a CAT tool in Easy Language translation. In Proceedings of the Second Workshop on Text Simplification, Accessibility and Readability, pages 1–10, Varna, Bulgaria. INCOMA Ltd., Shoumen, Bulgaria. URL: https://aclanthology.org/2023.tsar-1.1/. [↗](https://aclanthology.org/2023.tsar-1.1/)
4. Pierre Finnimore, Elisabeth Fritzsch, Daniel King, Alison Sneyd, Aneeq Ur Rehman, Fernando Alva-Manchego, and Andreas Vlachos. 2019. Strong Baselines for Complex Word Identification across Multiple Languages. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers), pages 970–977, Minneapolis, Minnesota, USA. Association for Computational Linguistics. URL: https://aclanthology.org/N19-1102, doi:10.18653/ v1/N19-1102. [↗](https://aclanthology.org/N19-1102/)
5. Sian Gooding. 2023. A Personalised Approach to Lexical Complexity. Doctoral Thesis, University of Cambridge, Cambridge, England. URL: https://doi.org/10.17863/CAM.116567. [↗](https://doi.org/10.17863/cam.116567)
6. Sian Gooding and Ekaterina Kochmar. 2019. Complex Word Identification as a Sequence Labelling Task. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 1148–1153, Florence, Italy. Association for Computational Linguistics. URL: https://aclanthology.org/P19-1109/, doi:10. 18653/v1/P19-1109. [↗](https://aclanthology.org/P19-1109/)
7. Sian Gooding, Ekaterina Kochmar, Seid Muhie Yimam, and Chris Biemann. 2021. Word Complexity is in the Eye of the Beholder. In Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 4439–4449, Online. Association for Computational Linguistics. URL: https://aclanthology.org/2021.naacl-main.351, doi: 10.18653/v1/2021.naacl-main.351. [↗](https://doi.org/10.18653/v1/2021.naacl-main.351)
8. Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2022. [LoRA: Low-Rank Adaptation of Large Language Models](https://md.research.elamur.ai/free/en/papers/hu-2021-lora-low-rank-adaptation-large/). In Proceedings of the International Conference on Learning Representations, Online. URL: [https://openreview.net/forum?id=nZeVKeeFYf9](https://openreview.net/forum?id=nZeVKeeFYf9). [↗](https://openreview.net/forum?id=nZeVKeeFYf9)
9. Tomoyuki Kajiwara and Mamoru Komachi. 2018. Complex Word Identification Based on Frequency in a Learner Corpus. In Proceedings of the Thirteenth Workshop on Innovative Use of NLP for Building Educational Applications, pages 195–199, New Orleans, Louisiana, USA. Association for Computational Linguistics. URL: https://aclanthology.org/W18-0521/, doi:10. 18653/v1/W18-0521. [↗](https://aclanthology.org/W18-0521/)
10. Percy Liang, Rishi Bommasani, Tony Lee, Dimitris Tsipras, Dilara Soylu, Michihiro Yasunaga, Yian Zhang, Deepak Narayanan, Yuhuai Wu, Ananya Kumar, Benjamin Newman, Binhang Yuan, Bobby Yan, Ce Zhang, Christian Cosgrove, Christopher D. Manning, Christopher Re, Diana Acosta-Navas, Drew A. Hudson, Eric Zelikman, Esin Durmus, Faisal Ladhak, Frieda Rong, Hongyu Ren, Huaxiu Yao, Jue Wang, Keshav Santhanam, Laurel Orr, Lucia Zheng, Mert Yuksekgonul, Mirac Suzgun, Nathan Kim, Neel Guha, Niladri S. Chatterji, Omar Khattab, Peter Henderson, Qian Huang, Ryan Andrew Chi, Sang Michael Xie, Shibani Santurkar, Surya Ganguli, Tatsunori Hashimoto, Thomas Icard, Tianyi Zhang, Vishrav Chaudhary, William Wang, Xuechen Li, Yifan Mai, Yuhui Zhang, and Yuta Koreeda. 2023. Holistic Evaluation of Language Models. Transactions on Machine Learning Research. URL: https://openreview.net/forum?id=iO4LZibEqW. [↗](https://doi.org/10.1111/nyas.15007)
11. Sourab Mangrulkar, Sylvain Gugger, Lysandre Debut, Younes Belkada, Sayak Paul, Benjamin Bossan, and Marian Tietz. 2022. PEFT: State-ofthe-art parameter-efficient fine-tuning methods. URL: [https://github.com/huggingface/peft](https://github.com/huggingface/peft). [↗](https://github.com/huggingface/peft)
12. Tomoya Mizumoto, Mamoru Komachi, Masaaki Nagata, and Yuji Matsumoto. 2011. Mining Revision Log of Language Learning SNS for Automated Japanese Error Correction of Second Language Learners. In Proceedings of 5th International Joint Conference on Natural Language Processing, pages 147–155, Chiang Mai, Thailand. Asian Federation of Natural Language Processing. [find](https://scholar.google.com/scholar?q=Tomoya+Mizumoto%2C+Mamoru+Komachi%2C+Masaaki+Nagata%2C+and+Yuji+Matsumoto%2E+2011%2E+Mining+Revision+Log+of+Language+Learning+SNS+for+Automated+Japanese+Error+Correction+of+Second+Language+Learners%2E+In+Proceedi)
13. Babak Naderi, Salar Mohtaj, Kaspar Ensikat, and Sebastian Möller. 2019. Subjective Assessment of Text Complexity: A Dataset for German Language. Version Number: 1. URL: https://arxiv.org/abs/1904.07733, doi: 10.48550/ARXIV.1904.07733. [↗](https://doi.org/10.48550/arxiv.1904.07733)
14. Jan Pfister and Andreas Hotho. 2024. Super-GLEBer: German Language Understanding Evaluation Benchmark. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), pages 7904–7923, Mexico City, Mexico. Association for Computational Linguistics. URL: https://aclanthology.org/2024.naacl-long.438/, doi:10.18653/v1/2024. naacl-long.438. [↗](https://doi.org/10.18653/v1/2024)
15. Jan Pfister, Julia Wunderle, and Andreas Hotho. 2025. [LLäMmlein: Transparent, Compact and Competitive German-Only Language Models from Scratch](https://md.research.elamur.ai/free/en/papers/pfister-2025-llammlein-transparent-compact-competitive-german/). In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 2227–2246, Vienna, Austria. Association for Computational Linguistics. URL: https://aclanthology.org/2025.acl-long.111/, doi: 10.18653/v1/2025.acl-long.111. [↗](https://doi.org/10.18653/v1/2025.acl-long.111)
16. Thorben Schomacker, Miriam Anschütz, Regina Stodden, Georg Groh, and Marina Tropmann-Frick. 2024. Overview of the GermEval 2024 Shared Task on Statement Segmentation in German Easy Language (StaGE). In Proceedings of GermEval 2024 Shared Task on Statement Segmentation in German Easy Language (StaGE), pages 1–14, Vienna, Austria. Association for Computational Lingustics. URL: https://aclanthology.org/2024.germeval-1.1. [↗](https://aclanthology.org/2024.germeval-1.1/)
17. Thorben Schomacker, Burak Tinman, Chris Biemann, and Marina Tropmann-Frick. 2026. LLMs for Easy Language Translation: A Case Study on German Public Authorities Web Pages. In KI 2025: Advances in Artificial Intelligence, pages 252–261, Cham. Springer Nature Switzerland. doi:10.1007/978-3-032-02813-6_20. [↗](https://doi.org/10.1007/978-3-032-02813-6_20)
18. Laura Seiffe, Fares Kallel, Sebastian Möller, Babak Naderi, and Roland Roller. 2022. Subjective Text Complexity Assessment for German. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 707–714, Marseille, France. European Language Resources Association. URL: https://aclanthology.org/2022.lrec-1.74. [↗](https://aclanthology.org/2022.lrec-1.74/)
19. Matthew Shardlow. 2013. [A Comparison of Techniques to Automatically Identify Complex Words](https://md.research.elamur.ai/free/en/papers/shardlow-2013-comparison-techniques-automatically-identify-complex/). In 51st Annual Meeting of the Association for Computational Linguistics Proceedings of the Student Research Workshop, pages 103–109, Sofia, Bulgaria. Association for Computational Linguistics. URL: https://aclanthology.org/P13-3015/. [↗](https://aclanthology.org/P13-3015/)
20. Matthew Shardlow. 2015. Lexical Simplification: Optimising the Pipeline. Dissertation, University of Manchester, Manchester, United Kingdom. URL: [https://pure.manchester.ac.uk/ws/portalfiles/portal/54575176/FULL_TEXT.PDF](https://pure.manchester.ac.uk/ws/portalfiles/portal/54575176/FULL_TEXT.PDF). [↗](https://pure.manchester.ac.uk/ws/portalfiles/portal/54575176/FULL_TEXT.PDF)
21. Matthew Shardlow, Fernando Alva-Manchego, Riza Batista-Navarro, Stefan Bott, Saul Calderon Ramirez, Rémi Cardon, Thomas François, Akio Hayakawa, Andrea Horbach, Anna Hülsing, Yusuke Ide, Joseph Marvin Imperial, Adam Nohejl, Kai North, Laura Occhipinti, Nelson Peréz Rojas, Nishat Raihan, Tharindu Ranasinghe, Martin Solis Salazar, Sanja Štajner, Marcos Zampieri, and Horacio Saggion. 2024. [find](https://scholar.google.com/scholar?q=Matthew+Shardlow%2C+Fernando+Alva%2DManchego%2C+Riza+Batista%2DNavarro%2C+Stefan+Bott%2C+Saul+Calderon+Ramirez%2C+R%C3%A9mi+Cardon%2C+Thomas+Fran%C3%A7ois%2C+Akio+Hayakawa%2C+Andrea+Horbach%2C+Anna+H%C3%BClsing%2C+Yusuke+Ide%2C+Joseph+Marvin)
22. The BEA 2024 Shared Task on the Multilingual Lexical Simplification Pipeline. In Proceedings of the 19th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2024), pages 571–589, Mexico City, Mexico. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=The+BEA+2024+Shared+Task+on+the+Multilingual+Lexical+Simplification+Pipeline%2E+In+Proceedings+of+the+19th+Workshop+on+Innovative+Use+of+NLP+for+Building+Educational+Applications+%28BEA+2024%29%2C+pages+571%E2%80%935)
23. Răzvan-Alexandru Smădu, David-Gabriel Ion, Dumitru-Clementin Cercel, Florin Pop, and Mihaela-Claudia Cercel. 2024. Investigating Large Language Models for Complex Word Identification in Multilingual and Multidomain Setups. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 16764–16800, Miami, Florida, USA. Association for Computational Linguistics. URL: https://aclanthology.org/2024.emnlp-main.933/, doi:10.18653/v1/2024. emnlp-main.933. [↗](https://doi.org/10.18653/v1/2024)
24. Teerapaun Tanprasert and David Kauchak. 2021. Flesch-Kincaid is Not a Text Simplification Evaluation Metric. In Proceedings of the 1st Workshop on Natural Language Generation, Evaluation, and Metrics (GEM 2021), pages 1–14, Online. Association for Computational Linguistics. URL: https://aclanthology.org/2021.gem-1.1, doi: 10.18653/v1/2021.gem-1.1. [↗](https://doi.org/10.18653/v1/2021.gem-1.1)
25. Alex Wang, Yada Pruksachatkun, Nikita Nangia, Amanpreet Singh, Julian Michael, Felix Hill, Omer Levy, and Samuel Bowman. 2019. [SuperGLUE: A Stickier Benchmark for General-Purpose Language Understanding Systems](https://md.research.elamur.ai/free/en/papers/wang-2019-superglue-stickier-benchmark-general-purpose/). In Advances in Neural Information Processing Systems, volume 32. Curran Associates, Inc. URL: [https://proceedings.neurips.cc/paper/2019/hash/](https://proceedings.neurips.cc/paper/2019/hash/) 4496bf24afe7fab6f046bf4923da8de6-Abstract. html. [↗](https://proceedings.neurips.cc/paper/2019/hash/)
26. Alex Wang, Amanpreet Singh, Julian Michael, Felix Hill, Omer Levy, and Samuel Bowman. 2018. [GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding](https://md.research.elamur.ai/free/en/papers/wang-2018-glue-multi-task-benchmark-analysis/). In Proceedings of the 2018 EMNLP Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP, pages 353–355, Brussels, Belgium. Association for Computational Linguistics. URL: http://aclweb.org/anthology/W18-5446, doi:10.18653/v1/W18-5446. [↗](https://doi.org/10.18653/v1/w18-5446)
27. Julia Wunderle, Anton Ehrmanntraut, Jan Pfister, Fotis Jannidis, and Andreas Hotho. 2025. [New Encoders for German Trained from Scratch: Comparing ModernGBERT with Converted LLM2Vec Models](https://md.research.elamur.ai/free/en/papers/wunderle-2026-new-encoders-german-trained-scratch/). ArXiv:2505.13136 [cs]. URL: http://arxiv.org/abs/2505.13136, doi: 10.48550/arXiv.2505.13136. [↗](https://doi.org/10.48550/arxiv.2505.13136)
28. Seid Muhie Yimam, Chris Biemann, Shervin Malmasi, Gustavo Paetzold, Lucia Specia, Sanja Štajner, Anaïs Tack, and Marcos Zampieri. 2018. A Report on the Complex Word Identification Shared Task 2018. In Proceedings of the Thirteenth Workshop on Innovative Use of NLP for Building Educational Applications, pages 66– 78, New Orleans, Louisiana, USA. Association for Computational Linguistics. URL: https://aclanthology.org/W18-0507/, doi:10.18653/ v1/W18-0507. [↗](https://aclanthology.org/W18-0507/)
29. Seid Muhie Yimam, Sanja Štajner, Martin Riedl, and Chris Biemann. 2017. Multilingual and Cross-Lingual Complex Word Identification. In Proceedings of the International Conference Recent Advances in Natural Language Processing, RANLP 2017, pages 813–822, Varna, Bulgaria. INCOMA Ltd. URL: https://aclanthology.org/R17-1104/, doi:10.26615/978-954-452-049-6_ 104. [↗](https://doi.org/10.26615/978-954-452-049-6_)

*Updated 1 November 2026*