---
title: "LangLingual — paper by Gupta et al. (2025) — summary and ideas"
description: "Teachers of English as a second language rarely have time to give every student individual practice and feedback, and… 27 references."
authors: ["Sammriddh Gupta", "Sonit Singh", "Aditya Joshi", "Mira Kim"]
year: 2025
venue: "arXiv:2510.23011"
source: "https://arxiv.org/abs/2510.23011"
doi: "10.48550/arxiv.2510.23011"
arxiv: "2510.23011"
licence: "CC BY-NC-SA 4.0 (arXiv)"
licence_url: "http://creativecommons.org/licenses/by-nc-sa/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/gupta-2025-langlingual-personalised-exercise-oriented-english/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/gupta-2025-langlingual-personalised-exercise-oriented-english/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2025](https://md.research.elamur.ai/free/en/papers/2025/page/18/) / LangLingual: A Personalised, Exercise-oriented English Language Learning Tool Leveraging Large Language Models

# LangLingual: A Personalised, Exercise-oriented English Language Learning Tool Leveraging Large Language Models

**Authors:** Sammriddh Gupta, Sonit Singh, [Aditya Joshi](https://md.research.elamur.ai/free/en/authors/aditya-joshi/) and Mira Kim · **Year:** 2025 · **Venue:** arXiv:2510.23011 · **Licence:** [CC BY-NC-SA 4.0 (arXiv)](http://creativecommons.org/licenses/by-nc-sa/4.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2510.23011) · [PDF](https://arxiv.org/pdf/2510.23011) · [DOI 10.48550/arxiv.2510.23011](https://doi.org/10.48550/arxiv.2510.23011) · [arXiv 2510.23011](https://arxiv.org/abs/2510.23011) · [OpenAlex](https://openalex.org/W7101842627)

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

## Summary
Teachers of English as a second language rarely have time to give every student individual practice and feedback, and general-purpose chatbots answer questions without tracking progress across sessions. The authors built LangLingual, a web-based conversational tutor on LangChain and OpenAI models that acts like a Socratic teacher, offering hints before answers. It rests on two curated resources. The first is a word bank of about 50,000 English words, each graded from level 1 to 14 by an experienced linguistics professor. The second is a CSV of learning resources organised by grammar and vocabulary area. A learner's level is estimated by combining a lemma lookup in the word bank with an LLM-as-judge prediction. The system detects when its own replies contain exercises and keeps track of the active one. It periodically mines the learner's accumulated messages for recurring weaknesses and recommends matching resources. A dashboard shows progress and session transcripts. Pre- and post-use surveys and three human-played learner personas indicated good usability, more motivation and perceived improvement. The work shows how a narrowly scoped, stateful LLM tutor differs from a generic chat assistant.

## Key points
- The Word Bank contains about 50,000 words in a `[word, level]` CSV, graded on a 1–14 scale (for example, `[ability, 1]`). It was supplied by a professor with 20 years of teaching experience.
- The Resources Dataset is a hand-curated CSV with the columns area, resource type, title, description, url and difficulty level. It covers topics such as articles, tenses, subject–verb agreement, phrasal verbs, idioms, punctuation and capitalisation.
- Proficiency is a weighted blend: 0.4 × word-bank level (mean or median over lemmatised matched words) plus 0.6 × an LLM-predicted level on the same 1–14 scale. The LLM part compensates for words missing from the list.
- Improvement-area analysis starts only after at least 3 learner turns. It concatenates all learner messages and asks the LLM for consistent error patterns rather than isolated slips.
- Evaluation consisted of a pre-survey, a 14-question post-survey (about 6.5 minutes on average) and persona-based testing with three personas:
  - Yuki, a Japanese student;
  - Sofia, a job-seeking accountant;
  - Charles, an Italian graphic designer.

## Interesting ideas
- Text difficulty is scored by lemmatising the input, looking each lemma up in a graded word list and taking the mean and median, then blending this with a model's estimate. The cheap, transparent signal anchors the opaque one.
- Weaknesses are diagnosed only after enough input has accumulated, and only patterns that recur across turns count. This avoids overreacting to a single noisy example.
- The tutor scans its own generated replies to decide whether they pose an exercise, and stores the active exercise as explicit state. This lets it tell whether the next user turn is an answer or a topic change.
- In persona-based evaluation, a human plays several scripted users with different levels, goals and contexts. This probes behaviour across user types without a large study.
- A fixed-weight combination of a rule-based score and an LLM judgment gives a simple, inspectable way to merge a deterministic witness with a model witness.

## For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

- **I1** (tokens): The lemmatiser could attach a graded-level feature to each lemma, giving an explainable readability score for any input text before parsing.
- **I2** (learn): Self-induced repair rules could be proposed only when the same parser error pattern repeats across many sentences, with one-off disagreements left as noise.
- **I3** (frames): Speech-act detection could mark utterances as open questions or tasks and keep a pending-obligation record. Later turns could then be interpreted as answers to it.
- **I4** (answer): Define reader personas (a young pupil, a non-native adult, an expert) and check that QA on fables and school tests adapts its explanations to each.
- **I5** (learn): LLM teacher tiers could be weighted against agreement among the witness parsers using explicit, logged weights, so every silver label's provenance stays explainable.

*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

- [UNSW Sydney](https://md.research.elamur.ai/free/en/universities/unsw-sydney/) — University, Sydney, Australia · 36 works · 1,509 citations

📚 Shelves (5): [🖥️ 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 · [🗄️ Information Systems](https://md.research.elamur.ai/free/en/shelves/computer-science/information-systems/) · 1,730 works

🧭 Topics (2): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,894 works · [English](https://md.research.elamur.ai/free/en/topics/english/) · 2,379 works

## Cite

Cite as: Sammriddh Gupta, Sonit Singh, Aditya Joshi and Mira Kim. 2025. LangLingual: A Personalised, Exercise-oriented English Language Learning Tool Leveraging Large Language Models. arXiv:2510.23011.

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

```bibtex
@misc{gupta2025langlingual,
  title = {{LangLingual: A Personalised, Exercise-oriented English Language Learning Tool Leveraging Large Language Models}},
  author = {Sammriddh Gupta, Sonit Singh, Aditya Joshi and Mira Kim},
  year = {2025},
  doi = {10.48550/arxiv.2510.23011},
  eprint = {2510.23011},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2510.23011}
}
```

- [← previous Lab17 @ Ahasis Shared Task 2025: Fine-Tuning and Prompting techniques for Sentiment Analysis of Saudi and Darija Dialects](https://md.research.elamur.ai/free/en/papers/hadhrami-2025-lab17-ahasis-shared-task-2025/)
- [next → Language Independent Named Entity Recognition via Orthogonal Transformation of Word Vectors](https://md.research.elamur.ai/free/en/papers/rakha-2025-language-independent-named-entity-recognition/)
- [↑ back to papers of 2025](https://md.research.elamur.ai/free/en/papers/2025/page/18/)

## References

27 references, 9 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Mohamed, A.M., Shaaban, T.S., Bakry, S.H., Guillén-Gámez, F.D., Strzelecki, A.: Empowering the faculty of education students: Applying AI’s potential for motivating and enhancing learning. Innov. High. Educ. (2024) [find](https://scholar.google.com/scholar?q=Mohamed%2C+A%2EM%2E%2C+Shaaban%2C+T%2ES%2E%2C+Bakry%2C+S%2EH%2E%2C+Guille%CC%81n%2DGa%CC%81mez%2C+F%2ED%2E%2C+Strzelecki%2C+A%2E%3A+Empowering+the+faculty+of+education+students%3A+Applying+AI%E2%80%99s+potential+for+motivating+and+enhancing+learning%2E+Innov%2E+Hi)
2. Wambsganss, T., Caines, A., Buttery, P.: [ALEN app: Argumentative writing support to foster English language learning](https://md.research.elamur.ai/free/en/papers/wambsganss-2022-alen-app-argumentative-writing-support/). In: Proceedings of the 17th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2022), pp. 134–140. Association for Computational Linguistics, Seattle, Washington (2022). https://doi.org/10.18653/v1/2022.bea-1.18 . https://aclanthology.org/2022.bea-1.18 [↗](https://doi.org/10.18653/v1/2022.bea-1.18)
3. Mahajan, M.: [BELA: Bot for English language acquisition](https://md.research.elamur.ai/free/en/papers/mahajan-2022-bela-bot-english-language-acquisition/). In: Proceedings of the Second Workshop on NLP for Positive Impact (NLP4PI), pp. 142–148. Association for Computational Linguistics, Abu Dhabi, United Arab Emirates (Hybrid) (2022). https://doi.org/10.18653/v1/2022.nlp4pi-1.17 . https://aclanthology.org/2022.nlp4pi-1.17 [↗](https://doi.org/10.18653/v1/2022.nlp4pi-1.17)
4. Papariello, L.: papluca/xlm-roberta-base-language-detection · Hugging Face (2024). [https://huggingface.co/papluca/xlm-roberta-base-language-detection](https://huggingface.co/papluca/xlm-roberta-base-language-detection) [↗](https://huggingface.co/papluca/xlm-roberta-base-language-detection)
5. Madnani, N., Burstein, J., Sabatini, J., Biggers, K., Andreyev, S.: [Language muse: Automated linguistic activity generation for English language learners](https://md.research.elamur.ai/free/en/papers/madnani-2016-language-muse-automated-linguistic-activity/). In: Proceedings of ACL-2016 System Demonstrations, pp. 79–84. Association for Computational Linguistics, Berlin, Germany (2016). https://doi.org/10.18653/v1/P16-4014 . https://aclanthology.org/P16-4014 [↗](https://doi.org/10.18653/v1/p16-4014)
6. Liang, K.-H., Davidson, S., Yuan, X., Panditharatne, S., Chen, C.-Y., Shea, R., Pham, D., Tan, Y., Voss, E., Fryer, L.: ChatBack: Investigating methods of providing grammatical error feedback in a GUI-based language learning chatbot. In: Proceedings of the 18th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2023), pp. 83–99. Association for Computational Linguistics, Toronto, Canada (2023). https://doi.org/10.18653/v1/2023.bea-1.7 . https://aclanthology.org/2023.bea-1.7 [↗](https://doi.org/10.18653/v1/2023.bea-1.7)
7. Tyen, G., Brenchley, M., Caines, A., Buttery, P.: Towards an open-domain chatbot for language practice. In: Proceedings of the 17th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2022), pp. 234–249. Association for Computational Linguistics, Seattle, Washington (2022). https://doi.org/10.18653/v1/2022.bea-1.28 . https://aclanthology.org/2022.bea-1.28 [↗](https://doi.org/10.18653/v1/2022.bea-1.28)
8. Europe, C.: Common European Framework of Reference for Languages: Learning, Teaching, Assessment (CEFR) - Common European Framework of Reference for Languages (CEFR) - www.coe.int (2001). [https://www.coe.int/en/web/common-european-framework-reference-languages](https://www.coe.int/en/web/common-european-framework-reference-languages) [↗](https://www.coe.int/en/web/common-european-framework-reference-languages)
9. Roller, S., Dinan, E., Goyal, N., Ju, D., Williamson, M., Liu, Y., Xu, J., Ott, M., Smith, E.M., Boureau, Y.-L., Weston, J.: [Recipes for building an open-domain chatbot](https://md.research.elamur.ai/free/en/papers/roller-2021-recipes-building-open-domain-chatbot/). In: Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pp. 300–325. Association for Computational Linguistics, Online (2021). https://doi.org/10.18653/v1/2021.eacl-main.24 . https://aclanthology.org/2021.eacl-main.24 [↗](https://doi.org/10.18653/v1/2021.eacl-main.24)
10. Fan, A., Lewis, M., Dauphin, Y.: [Hierarchical neural story generation](https://md.research.elamur.ai/free/en/papers/fan-2018-hierarchical-neural-story-generation/). In: Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pp. 889–898. Association for Computational Linguistics, Melbourne, Australia (2018). https://doi.org/10.18653/v1/P18-1082 . https://aclanthology.org/P18-1082 [↗](https://doi.org/10.18653/v1/p18-1082)
11. Sennrich, R., Haddow, B., Birch, A.: [Neural machine translation of rare words with subword units](https://md.research.elamur.ai/free/en/papers/sennrich-2016-neural-machine-translation-rare-words/). In: Erk, K., Smith, N.A. (eds.) Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pp. 1715–1725. Association for Computational Linguistics, Berlin, Germany (2016). https://doi.org/10.18653/v1/P16-1162 . https://aclanthology.org/P16-1162 [↗](https://doi.org/10.18653/v1/p16-1162)
12. Shatz, I.: Refining and modifying the EFCAMDAT: Lessons from creating a new corpus from an existing large-scale English learner language database. https://www.jbe-platform.com/content/journals/10.1075/ijlcr.20009.sha [↗](https://doi.org/10.1075/ijlcr.20009.sha)
13. Tack, A., Kochmar, E., Yuan, Z., Bibauw, S., Piech, C.: The BEA 2023 shared task on generating AI teacher responses in educational dialogues. In: Proceedings of the 18th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2023), pp. 785–795. Association for Computational Linguistics, Toronto, Canada (2023). https://doi.org/10.18653/v1/2023.bea-1.64 . https://aclanthology.org/2023.bea-1.64 [↗](https://doi.org/10.18653/v1/2023.bea-1.64)
14. Radford, A., Wu, J., Child, R., Luan, D., Amodei, D., Sutskever, I.: Language Models are Unsupervised Multitask Learners (2019) [find](https://scholar.google.com/scholar?q=Radford%2C+A%2E%2C+Wu%2C+J%2E%2C+Child%2C+R%2E%2C+Luan%2C+D%2E%2C+Amodei%2C+D%2E%2C+Sutskever%2C+I%2E%3A+Language+Models+are+Unsupervised+Multitask+Learners+%282019%29)
15. Brown, T.B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., Agarwal, S., Herbert-Voss, A., Krueger, G., Henighan, T., Child, R., Ramesh, A., Ziegler, D.M., Wu, J., Winter, C., Hesse, C., Chen, M., Sigler, E., Litwin, M., Gray, S., Chess, B., Clark, J., Berner, C., McCandlish, S., Radford, A., Sutskever, I., Amodei, D.: [Language Models are Few-Shot Learners](https://md.research.elamur.ai/free/en/papers/brown-2020-language-models-few-shot-learners/) (2020). https://arxiv.org/abs/2005.14165 [↗](https://arxiv.org/abs/2005.14165)
16. Reimers, N., Gurevych, I.: [Sentence-BERT: Sentence embeddings using Siamese BERT-networks](https://md.research.elamur.ai/free/en/papers/reimers-2019-sentence-bert-sentence-embeddings-siamese/). 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), pp. 3982–3992. Association for Computational Linguistics, Hong Kong, China (2019). https://doi.org/10.18653/v1/D19-1410 . https://aclanthology.org/D19-1410 [↗](https://doi.org/10.18653/v1/d19-1410)
17. Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., Küttler, H., Lewis, M., Yih, W.-t., Rocktäschel, T., Riedel, S., Kiela, D.: [Retrieval-augmented generation for knowledge-intensive nlp tasks](https://md.research.elamur.ai/free/en/papers/lewis-2020-retrieval-augmented-generation-knowledge-intensive/). In: Proceedings of the 34th International Conference on Neural Information Processing Systems. NIPS ’20. Curran Associates Inc., Red Hook, NY, USA (2020)
18. Xu, A., Monroe, W., Bicknell, K.: Large language model augmented exercise retrieval for personalized language learning. In: Proceedings of the 14th Learning Analytics and Knowledge Conference. LAK ’24, pp. 284–294. Association for Computing Machinery, New York, NY, USA (2024). https://doi.org/10.1145/3636555.3636883 . https://doi.org/10.1145/3636555.3636883 [↗](https://doi.org/10.1145/3636555.3636883)
19. Kohnke, L., Zou, D., Su, F.: Exploring the potential of genai for personalised english teaching: Learners’ experiences and perceptions. Computers and Education: Artificial Intelligence 8 , 100371 (2025) https://doi.org/10.1016/j.caeai.2025.100371 [↗](https://doi.org/10.1016/j.caeai.2025.100371)
20. Cui, P., Sachan, M.: Adaptive and personalized exercise generation for online language learning. In: Rogers, A., Boyd-Graber, J., Okazaki, N. (eds.) Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pp. 10184–10198. Association for Computational Linguistics, Toronto, Canada (2023). https://doi.org/10.18653/v1/2023.acl-long.567 . https://aclanthology.org/2023.acl-long.567/ [↗](https://doi.org/10.18653/v1/2023.acl-long.567)
21. Kim, M., Nallbani, A.L., Stovall, A.R.: Exploring llm-based chatbot for language learning and cultivation of growth mindset. In: Extended Abstracts of the CHI Conference on Human Factors in Computing Systems. CHI EA ’24. Association for Computing Machinery, New York, NY, USA (2024). https://doi.org/10.1145/3613905.3648628 . https://doi.org/10.1145/3613905.3648628 [↗](https://doi.org/10.1145/3613905.3648628)
22. Yadegaridehkordi, E., Foroughi, B., Ghobakhloo, M.: Factors affecting academic staff’s willingness to use ChatGPT for teaching and learning: A PLS-SEM and ANN approach. Innov. High. Educ. (2025) [find](https://scholar.google.com/scholar?q=Yadegaridehkordi%2C+E%2E%2C+Foroughi%2C+B%2E%2C+Ghobakhloo%2C+M%2E%3A+Factors+affecting+academic+staff%E2%80%99s+willingness+to+use+ChatGPT+for+teaching+and+learning%3A+A+PLS%2DSEM+and+ANN+approach%2E+Innov%2E+High%2E+Educ%2E+%282025%29)
23. Chase, H.: LangChain (2022). [https://python.langchain.com/docs/introduction/](https://python.langchain.com/docs/introduction/) [↗](https://python.langchain.com/docs/introduction/)
24. Zanudin, N.N., Sulaiman, S., Samingan, M.R., Mohamed, H., Raof, S.K.S.A., Abd Samad, A.R.: Case study on prototyping educational applications using persona-based approach. In: 2021 8th International Conference on Computer and Communication Engineering (ICCCE), pp. 93–98 (2021). IEEE [find](https://scholar.google.com/scholar?q=Zanudin%2C+N%2EN%2E%2C+Sulaiman%2C+S%2E%2C+Samingan%2C+M%2ER%2E%2C+Mohamed%2C+H%2E%2C+Raof%2C+S%2EK%2ES%2EA%2E%2C+Abd+Samad%2C+A%2ER%2E%3A+Case+study+on+prototyping+educational+applications+using+persona%2Dbased+approach%2E+In%3A+2021+8th+International+C)
25. Jandaghi, P., Sheng, X., Bai, X., Pujara, J., Sidahmed, H.: Faithful persona-based conversational dataset generation with large language models. In: Findings of the Association for Computational Linguistics ACL 2024, pp. 15245–15270 (2024) [find](https://scholar.google.com/scholar?q=Jandaghi%2C+P%2E%2C+Sheng%2C+X%2E%2C+Bai%2C+X%2E%2C+Pujara%2C+J%2E%2C+Sidahmed%2C+H%2E%3A+Faithful+persona%2Dbased+conversational+dataset+generation+with+large+language+models%2E+In%3A+Findings+of+the+Association+for+Computational+Lingu)
26. Henka, A., Zimmermann, G.: Persona based accessibility testing: Towards user-centered accessibility evaluation. In: International Conference on Human-Computer Interaction, pp. 226–231 (2014). Springer [find](https://scholar.google.com/scholar?q=Henka%2C+A%2E%2C+Zimmermann%2C+G%2E%3A+Persona+based+accessibility+testing%3A+Towards+user%2Dcentered+accessibility+evaluation%2E+In%3A+International+Conference+on+Human%2DComputer+Interaction%2C+pp%2E+226%E2%80%93231+%282014%29%2E+Springer)
27. Stojmenova, E., Lugmayr, A., Dinevski, D.: Persona-based expert review of an elearning system for adults. In: 2013 IEEE International Conference on Multimedia and Expo Workshops (ICMEW), pp. 1–4 (2013). IEEE [find](https://scholar.google.com/scholar?q=Stojmenova%2C+E%2E%2C+Lugmayr%2C+A%2E%2C+Dinevski%2C+D%2E%3A+Persona%2Dbased+expert+review+of+an+elearning+system+for+adults%2E+In%3A+2013+IEEE+International+Conference+on+Multimedia+and+Expo+Workshops+%28ICMEW%29%2C+pp%2E+1%E2%80%934+%282013)

*Updated 31 October 2026*