---
title: "Fine-tuning Transformers with Additional Context to Classify"
description: "Teachers use \"talk moves\", discourse strategies from accountable talk theory, to get… 78.92% F1. 42 references, cited by 16 papers here."
authors: ["Abhijit Suresh", "Jennifer Jacobs", "Margaret Perkoff", "James H. Martin", "Tamara Sumner"]
year: 2022
venue: "Proceedings of the 17th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2022)"
source: "https://aclanthology.org/2022.bea-1.11/"
doi: "10.18653/v1/2022.bea-1.11"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/suresh-2022-fine-tuning-transformers-additional-context/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/suresh-2022-fine-tuning-transformers-additional-context/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2022](https://md.research.elamur.ai/free/en/papers/2022/page/5/) / Fine-tuning Transformers with Additional Context to Classify Discursive Moves in Mathematics Classrooms

# Fine-tuning Transformers with Additional Context to Classify Discursive Moves in Mathematics Classrooms

**Authors:** [Abhijit Suresh](https://md.research.elamur.ai/free/en/authors/abhijit-suresh/), [Jennifer Jacobs](https://md.research.elamur.ai/free/en/authors/jennifer-jacobs/), [Margaret Perkoff](https://md.research.elamur.ai/free/en/authors/margaret-perkoff/), [James H. Martin](https://md.research.elamur.ai/free/en/authors/james-h-martin/) and [Tamara Sumner](https://md.research.elamur.ai/free/en/authors/tamara-sumner/) · **Year:** 2022 · **Venue:** [Proceedings of the 17th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2022)](https://md.research.elamur.ai/free/en/venues/bea/) ([BEA 2022](https://md.research.elamur.ai/free/en/venues/bea/2022/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2022.bea-1.11/) · [PDF](https://aclanthology.org/2022.bea-1.11.pdf) · [DOI 10.18653/v1/2022.bea-1.11](https://doi.org/10.18653/v1/2022.bea-1.11) · [OpenAlex](https://openalex.org/W4287855004)

**Citations:** 30 citations (OpenAlex, 2026-10-14) · [16](https://md.research.elamur.ai/free/en/papers/suresh-2022-fine-tuning-transformers-additional-context/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
Teachers use "talk moves", discourse strategies from accountable talk theory, to get students explaining and building on each other's ideas. Coding these moves in lesson transcripts by hand is slow and needs trained experts. This work tries to make automatic classification more accurate by giving the model more of the surrounding conversation. Earlier systems fed Roberta only the teacher utterance and the student turn just before it. The authors instead built inputs that add several preceding turns, several following turns, or both, borrowing the idea from spoken dialogue systems. They tested this on the TalkMoves dataset of K-12 mathematics lessons. They also compared long-sequence architectures (Longformer, BigBird), XLNet and the math-oriented MathBERT. A balanced window of seven earlier and seven later utterances worked best with Roberta-Base, reaching a new state of the art on this benchmark. Neither the long-context models nor the math pretraining helped. The results matter for tools that give teachers automated feedback on their classroom discourse.

## Key points
- **Data:** the TalkMoves dataset has 567 human-transcribed K-12 math lesson transcripts, with 174,186 teacher and 59,874 student utterances. Each teacher utterance is labeled as one of six talk moves or "None", and the classes are highly imbalanced.
- **Best result:** Roberta-Base with Previous 7 – Subsequent 7 context scored 78.92% F1 (MCC 0.7805). The earlier immediate-student baseline scored 76.32% F1, a gain of about 2.6 points. With no context at all, the model reached only 71.93% F1.
- **Context direction:** using earlier turns alone peaked at 77.63% F1 (Previous 15). One following turn alone gave 74.16% F1. Combining both directions consistently did better.
- **Other models** (same 7+7 window):
  - Longformer: 78.47% F1
  - XLNet: 78.06% F1
  - BigBird: 77.89% F1
  - MathBERT: 70.18% F1
  
  None of them beat Roberta-Base.
- **Error analysis:** confusion matrices show the wider window raised F1 for almost every talk move except Restating. "Getting students to relate" and "Revoicing" remained the hardest classes.

*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 Colorado Boulder](https://md.research.elamur.ai/free/en/universities/university-of-colorado-boulder/) — University, Boulder, United States · 76 works · 51,396 citations
- [University of Colorado System](https://md.research.elamur.ai/free/en/universities/university-of-colorado-system/) — University, Boulder, United States · 11 works · 1,747 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 19,503 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 18,805 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 17,959 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 17,789 works

## Cite

Cite as: Abhijit Suresh, Jennifer Jacobs, Margaret Perkoff, James H. Martin and Tamara Sumner. 2022. Fine-tuning Transformers with Additional Context to Classify Discursive Moves in Mathematics Classrooms. Proceedings of the 17th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2022).

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

```bibtex
@inproceedings{suresh2022fine,
  title = {{Fine-tuning Transformers with Additional Context to Classify Discursive Moves in Mathematics Classrooms}},
  author = {Abhijit Suresh, Jennifer Jacobs, Margaret Perkoff, James H. Martin and Tamara Sumner},
  booktitle = {Proceedings of the 17th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2022)},
  year = {2022},
  doi = {10.18653/v1/2022.bea-1.11},
  url = {https://doi.org/10.18653/v1/2022.bea-1.11}
}
```

- [← previous Fact Checking with Insufficient Evidence](https://md.research.elamur.ai/free/en/papers/atanasova-2022-fact-checking-insufficient-evidence/)
- [next → F-coref: Fast, Accurate and Easy to Use Coreference Resolution](https://md.research.elamur.ai/free/en/papers/otmazgin-2022-f-coref-fast-accurate-easy/)
- [↑ back to papers of 2022](https://md.research.elamur.ai/free/en/papers/2022/page/5/)

## References

42 references, 12 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. David Adamson, Colin Ashe, Hyeju Jang, David Yaron, and Carolyn Penstein Rosé. 2013. Intensification of group knowledge exchange with academically productive talk agents. In CSCL (1), pages 10–17. [find](https://scholar.google.com/scholar?q=David+Adamson%2C+Colin+Ashe%2C+Hyeju+Jang%2C+David+Yaron%2C+and+Carolyn+Penstein+Ros%C3%A9%2E+2013%2E+Intensification+of+group+knowledge+exchange+with+academically+productive+talk+agents%2E+In+CSCL+%281%29%2C+pages+10%E2%80%9317%2E)
2. Zihang Dai, Zhilin Yang, Yiming Yang, Jaime Carbonell, Quoc V Le, and Ruslan Salakhutdinov. 2019. Transformer-xl: Attentive language models beyond a fixed-length context. arXiv preprint arXiv:1901.02860. [find](https://scholar.google.com/scholar?q=Zihang+Dai%2C+Zhilin+Yang%2C+Yiming+Yang%2C+Jaime+Carbonell%2C+Quoc+V+Le%2C+and+Ruslan+Salakhutdinov%2E+2019%2E+Transformer%2Dxl%3A+Attentive+language+models+beyond+a+fixed%2Dlength+context%2E+arXiv+preprint+arXiv%3A1901%2E028)
3. Iz Beltagy, Matthew E Peters, and Arman Cohan. 2020. [Longformer: The long-document transformer](https://md.research.elamur.ai/free/en/papers/beltagy-2020-longformer-long-document-transformer/). arXiv preprint arXiv:2004.05150.
4. Melissa Boston. 2012. Assessing instructional quality in mathematics. The Elementary School Journal, 113(1):76–104. [find](https://scholar.google.com/scholar?q=Melissa+Boston%2E+2012%2E+Assessing+instructional+quality+in+mathematics%2E+The+Elementary+School+Journal%2C+113%281%29%3A76%E2%80%93104%2E)
5. Meghan E Dale, Amanda J Godley, Sarah A Capello, Patrick J Donnelly, Sidney K D’Mello, and Sean P Kelly. 2022. Toward the automated analysis of teacher talk in secondary ela classrooms. Teaching and Teacher Education, 110:103584. [find](https://scholar.google.com/scholar?q=Meghan+E+Dale%2C+Amanda+J+Godley%2C+Sarah+A+Capello%2C+Patrick+J+Donnelly%2C+Sidney+K+D%E2%80%99Mello%2C+and+Sean+P+Kelly%2E+2022%2E+Toward+the+automated+analysis+of+teacher+talk+in+secondary+ela+classrooms%2E+Teaching+and+T)
6. Amber G Candela, Melissa D Boston, and Juli K Dixon. 2020. Discourse actions to promote student access. [find](https://scholar.google.com/scholar?q=Amber+G+Candela%2C+Melissa+D+Boston%2C+and+Juli+K+Dixon%2E+2020%2E+Discourse+actions+to+promote+student+access%2E)
7. Dorottya Demszky, Jing Liu, Heather C Hill, Dan Jurafsky, and Chris Piech. 2021. Can automated feedback improve teachers’ uptake of student ideas? evidence from a randomized controlled trial in a large-scale online course. [find](https://scholar.google.com/scholar?q=Dorottya+Demszky%2C+Jing+Liu%2C+Heather+C+Hill%2C+Dan+Jurafsky%2C+and+Chris+Piech%2E+2021%2E+Can+automated+feedback+improve+teachers%E2%80%99+uptake+of+student+ideas%3F+evidence+from+a+randomized+controlled+trial+in+a+larg)
8. Sarah Michaels, Mary Catherine O’Connor, Megan Williams Hall, and Lauren B Resnick. 2010. Accountable talk® sourcebook. Pittsburg, PA: Institute for Learning University of Pittsburgh. Murphy, PK, Wilkinson, IAG, Soter, AO, Hennessey, MN, & Alexander, JF. [find](https://scholar.google.com/scholar?q=Sarah+Michaels%2C+Mary+Catherine+O%E2%80%99Connor%2C+Megan+Williams+Hall%2C+and+Lauren+B+Resnick%2E+2010%2E+Accountable+talk%C2%AE+sourcebook%2E+Pittsburg%2C+PA%3A+Institute+for+Learning+University+of+Pittsburgh%2E+Murphy%2C+PK%2C+Wilk)
9. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2018. [Bert: Pre-training of deep bidirectional transformers for language understanding](https://md.research.elamur.ai/free/en/papers/devlin-2019-bert-pre-training-deep-bidirectional/). arXiv preprint arXiv:1810.04805.
10. Catherine O’Connor and Sarah Michaels. 2019. Supporting teachers in taking up productive talk moves: The long road to professional learning at scale. International Journal of Educational Research, 97:166– 175. [find](https://scholar.google.com/scholar?q=Catherine+O%E2%80%99Connor+and+Sarah+Michaels%2E+2019%2E+Supporting+teachers+in+taking+up+productive+talk+moves%3A+The+long+road+to+professional+learning+at+scale%2E+International+Journal+of+Educational+Research%2C+97%3A)
11. Patrick J Donnelly, Nathaniel Blanchard, Andrew M Olney, Sean Kelly, Martin Nystrand, and Sidney K D’Mello. 2017. Words matter: automatic detection of teacher questions in live classroom discourse using linguistics, acoustics, and context. In Proceedings of the Seventh International Learning Analyti… [find](https://scholar.google.com/scholar?q=Patrick+J+Donnelly%2C+Nathaniel+Blanchard%2C+Andrew+M+Olney%2C+Sean+Kelly%2C+Martin+Nystrand%2C+and+Sidney+K+D%E2%80%99Mello%2E+2017%2E+Words+matter%3A+automatic+detection+of+teacher+questions+in+live+classroom+discourse+usi)
12. Catherine O’Connor, Sarah Michaels, and Suzanne Chapin. 2015. Scaling down” to explore the role of talk in learning: From district intervention to controlled classroom study. Socializing intelligence through academic talk and dialogue, pages 111–126. [find](https://scholar.google.com/scholar?q=Catherine+O%E2%80%99Connor%2C+Sarah+Michaels%2C+and+Suzanne+Chapin%2E+2015%2E+Scaling+down%E2%80%9D+to+explore+the+role+of+talk+in+learning%3A+From+district+intervention+to+controlled+classroom+study%2E+Socializing+intelligence+)
13. Beth A Herbel-Eisenmann. 2017. Mathematics Discourse in Secondary Classrooms: A Practice-based Resource for Professional Learning: Facilitator Guide. Math Solutions. [find](https://scholar.google.com/scholar?q=Beth+A+Herbel%2DEisenmann%2E+2017%2E+Mathematics+Discourse+in+Secondary+Classrooms%3A+A+Practice%2Dbased+Resource+for+Professional+Learning%3A+Facilitator+Guide%2E+Math+Solutions%2E)
14. Shuai Peng, Ke Yuan, Liangcai Gao, and Zhi Tang. 2021. Mathbert: A pre-trained model for mathematical formula understanding. arXiv preprint arXiv:2105.00377. — [arXiv:2105.00377](https://arxiv.org/abs/2105.00377) [find](https://scholar.google.com/scholar?q=Shuai+Peng%2C+Ke+Yuan%2C+Liangcai+Gao%2C+and+Zhi+Tang%2E+2021%2E+Mathbert%3A+A+pre%2Dtrained+model+for+mathematical+formula+understanding%2E+arXiv+preprint+arXiv%3A2105%2E00377%2E+%E2%80%94+%5BarXiv%3A2105%2E00377%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs)
15. Cindy E Hmelo-Silver, Clark A Chinn, Angela M O’Donnell, and Carol Chan. 2013. The international handbook of collaborative learning. [find](https://scholar.google.com/scholar?q=Cindy+E+Hmelo%2DSilver%2C+Clark+A+Chinn%2C+Angela+M+O%E2%80%99Donnell%2C+and+Carol+Chan%2E+2013%2E+The+international+handbook+of+collaborative+learning%2E)
16. Samuel L Pugh, Shree Krishna Subburaj, Arjun Ramesh Rao, Angela EB Stewart, Jessica Andrews-Todd, and Sidney K D’Mello. 2021. Say what? automatic modeling of collaborative problem solving skills from student speech in the wild. International Educational Data Mining Society. [find](https://scholar.google.com/scholar?q=Samuel+L+Pugh%2C+Shree+Krishna+Subburaj%2C+Arjun+Ramesh+Rao%2C+Angela+EB+Stewart%2C+Jessica+Andrews%2DTodd%2C+and+Sidney+K+D%E2%80%99Mello%2E+2021%2E+Say+what%3F+automatic+modeling+of+collaborative+problem+solving+skills+from+)
17. Jennifer Jacobs, Karla Scornavacco, Charis Harty, Abhijit Suresh, Vivian Lai, and Tamara Sumner. 2022. Promoting rich discussions in mathematics classrooms: Using personalized, automated feedback to support reflection and instructional change. Teaching and Teacher Education. [find](https://scholar.google.com/scholar?q=Jennifer+Jacobs%2C+Karla+Scornavacco%2C+Charis+Harty%2C+Abhijit+Suresh%2C+Vivian+Lai%2C+and+Tamara+Sumner%2E+2022%2E+Promoting+rich+discussions+in+mathematics+classrooms%3A+Using+personalized%2C+automated+feedback+to+s)
18. Lauren B Resnick, Christa SC Asterhan, and Sherice N Clarke. 2018. Accountable talk: Instructional dialogue that builds the mind. Geneva, Switzerland: The International Academy of Education (IAE) and the International Bureau of Education (IBE) of the United Nations Educational, Scientific and Cultur… [find](https://scholar.google.com/scholar?q=Lauren+B+Resnick%2C+Christa+SC+Asterhan%2C+and+Sherice+N+Clarke%2E+2018%2E+Accountable+talk%3A+Instructional+dialogue+that+builds+the+mind%2E+Geneva%2C+Switzerland%3A+The+International+Academy+of+Education+%28IAE%29+and+)
19. Emily Jensen, Meghan Dale, Patrick J Donnelly, Cath-lyn Stone, Sean Kelly, Amanda Godley, and Sidney K D’Mello. 2020. Toward automated feedback on teacher discourse to enhance teacher learning. In Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems, pages 1–13. [find](https://scholar.google.com/scholar?q=Emily+Jensen%2C+Meghan+Dale%2C+Patrick+J+Donnelly%2C+Cath%2Dlyn+Stone%2C+Sean+Kelly%2C+Amanda+Godley%2C+and+Sidney+K+D%E2%80%99Mello%2E+2020%2E+Toward+automated+feedback+on+teacher+discourse+to+enhance+teacher+learning%2E+In+Pro)
20. Christopher Schröder, Andreas Niekler, and Martin Potthast. 2021. Uncertainty-based query strategies for active learning with transformers. arXiv preprint arXiv:2107.05687. — [arXiv:2107.05687](https://arxiv.org/abs/2107.05687) [find](https://scholar.google.com/scholar?q=Christopher+Schr%C3%B6der%2C+Andreas+Niekler%2C+and+Martin+Potthast%2E+2021%2E+Uncertainty%2Dbased+query+strategies+for+active+learning+with+transformers%2E+arXiv+preprint+arXiv%3A2107%2E05687%2E+%E2%80%94+%5BarXiv%3A2107%2E05687%5D%28https%3A)
21. Mandar Joshi, Omer Levy, Daniel S Weld, and Luke Zettlemoyer. 2019. [Bert for coreference resolution: Baselines and analysis](https://md.research.elamur.ai/free/en/papers/joshi-2019-bert-coreference-resolution-baselines-analysis/). arXiv preprint arXiv:1908.09091.
22. E Schukat-Talamazzini, T Kuhn, and H Niemann. 1994. Speech recognition for spoken dialogue systems. In Progress and Prospects of Speech Research and Technology: Proc. of the CRIM/FORWISS Workshop, PAI, volume 1, pages 110–120. [find](https://scholar.google.com/scholar?q=E+Schukat%2DTalamazzini%2C+T+Kuhn%2C+and+H+Niemann%2E+1994%2E+Speech+recognition+for+spoken+dialogue+systems%2E+In+Progress+and+Prospects+of+Speech+Research+and+Technology%3A+Proc%2E+of+the+CRIM%2FFORWISS+Workshop%2C+PAI)
23. Olga Kovaleva, Alexey Romanov, Anna Rogers, and Anna Rumshisky. 2019. [Revealing the dark secrets of bert](https://md.research.elamur.ai/free/en/papers/kovaleva-2019-revealing-dark-secrets-bert/). arXiv preprint arXiv:1908.08593.
24. Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. [Roberta: A robustly optimized bert pretraining approach](https://md.research.elamur.ai/free/en/papers/liu-2019-roberta-robustly-optimized-bert-pretraining/). arXiv preprint arXiv:1907.11692.
25. Burr Settles. 2009. [Active learning literature survey](https://md.research.elamur.ai/free/en/papers/settles-2009-active-learning-literature-survey/).
26. Yu Song, Shunwei Lei, Tianyong Hao, Zixin Lan, and Ying Ding. 2021. Automatic classification of semantic content of classroom dialogue. Journal of Educational Computing Research, 59(3):496–521. [find](https://scholar.google.com/scholar?q=Yu+Song%2C+Shunwei+Lei%2C+Tianyong+Hao%2C+Zixin+Lan%2C+and+Ying+Ding%2E+2021%2E+Automatic+classification+of+semantic+content+of+classroom+dialogue%2E+Journal+of+Educational+Computing+Research%2C+59%283%29%3A496%E2%80%93521%2E)
27. Hugh Mehan. 1979. Learning lessons. Harvard University Press Cambridge, MA. [find](https://scholar.google.com/scholar?q=Hugh+Mehan%2E+1979%2E+Learning+lessons%2E+Harvard+University+Press+Cambridge%2C+MA%2E)
28. Abhijit Suresh, Jennifer Jacobs, Charis Clevenger, Vivian Lai, Chenhao Tan, James H Martin, and Tamara Sumner. 2021a. Using ai to promote equitable classroom discussions: The talkmoves application. In International Conference on Artificial Intelligence in Education, pages 344–348. Springer. [find](https://scholar.google.com/scholar?q=Abhijit+Suresh%2C+Jennifer+Jacobs%2C+Charis+Clevenger%2C+Vivian+Lai%2C+Chenhao+Tan%2C+James+H+Martin%2C+and+Tamara+Sumner%2E+2021a%2E+Using+ai+to+promote+equitable+classroom+discussions%3A+The+talkmoves+application%2E+In)
29. Sarah Michaels and Catherine O’Connor. 2015. Conceptualizing talk moves as tools: Professional development approaches for academically productive discussion. Socializing intelligence through talk and dialogue, pages 347–362. [find](https://scholar.google.com/scholar?q=Sarah+Michaels+and+Catherine+O%E2%80%99Connor%2E+2015%2E+Conceptualizing+talk+moves+as+tools%3A+Professional+development+approaches+for+academically+productive+discussion%2E+Socializing+intelligence+through+talk+and+)
30. Abhijit Suresh, Jennifer Jacobs, Charis Harty, Margaret Perkoff, James H Martin, and Tamara Sumner. 2022. [The talkmoves dataset: K-12 mathematics lesson transcripts annotated for teacher and student discursive moves](https://md.research.elamur.ai/free/en/papers/suresh-2022-talkmoves-dataset-k-12-mathematics/). 13th International Conference on Language Resources and Evaluation (LREC 2022).
31. Manzil Zaheer, Guru Guruganesh, Kumar Avinava Dubey, Joshua Ainslie, Chris Alberti, Santiago Ontanon, Philip Pham, Anirudh Ravula, Qifan Wang, Li Yang, et al. 2020. [Big bird: Transformers for longer sequences](https://md.research.elamur.ai/free/en/papers/zaheer-2020-big-bird-transformers-longer-sequences/). Advances in Neural Information Processing Systems, 33:17283–17297.
32. Abhijit Suresh, Jennifer Jacobs, Vivian Lai, Chenhao Tan, Wayne Ward, James H Martin, and Tamara Sumner. 2021b. Using transformers to provide teachers with personalized feedback on their classroom discourse: The talkmoves application. [find](https://scholar.google.com/scholar?q=Abhijit+Suresh%2C+Jennifer+Jacobs%2C+Vivian+Lai%2C+Chenhao+Tan%2C+Wayne+Ward%2C+James+H+Martin%2C+and+Tamara+Sumner%2E+2021b%2E+Using+transformers+to+provide+teachers+with+personalized+feedback+on+their+classroom+dis)
33. AAAI 2021 Spring Symposium on Artificial Intelligence for K-12 Education. [find](https://scholar.google.com/scholar?q=AAAI+2021+Spring+Symposium+on+Artificial+Intelligence+for+K%2D12+Education%2E)
34. Abhijit Suresh, Tamara Sumner, Jennifer Jacobs, Bill Foland, and Wayne Ward. 2019. Automating analysis and feedback to improve mathematics teachers’ classroom discourse. In Proceedings of the AAAI conference on artificial intelligence, volume 33, pages 9721–9728. [find](https://scholar.google.com/scholar?q=Abhijit+Suresh%2C+Tamara+Sumner%2C+Jennifer+Jacobs%2C+Bill+Foland%2C+and+Wayne+Ward%2E+2019%2E+Automating+analysis+and+feedback+to+improve+mathematics+teachers%E2%80%99+classroom+discourse%2E+In+Proceedings+of+the+AAAI+con)
35. Stergios Tegos, Stavros Demetriadis, and Anastasios Karakostas. 2015. Promoting academically productive talk with conversational agent interventions in collaborative learning settings. Computers & Education, 87:309–325. [find](https://scholar.google.com/scholar?q=Stergios+Tegos%2C+Stavros+Demetriadis%2C+and+Anastasios+Karakostas%2E+2015%2E+Promoting+academically+productive+talk+with+conversational+agent+interventions+in+collaborative+learning+settings%2E+Computers+%26+Edu)
36. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. [Attention is all you need](https://md.research.elamur.ai/free/en/papers/vaswani-2017-attention-all-need/). Advances in neural information processing systems, 30.
37. Lev Vygotsky. 1978. Interaction between learning and development. Readings on the development of children, 23(3):34–41. [find](https://scholar.google.com/scholar?q=Lev+Vygotsky%2E+1978%2E+Interaction+between+learning+and+development%2E+Readings+on+the+development+of+children%2C+23%283%29%3A34%E2%80%9341%2E)
38. Noreen M Webb, Megan L Franke, Marsha Ing, Angela Chan, Tondra De, Deanna Freund, and Dan Battey. 2008. The role of teacher instructional practices in student collaboration. Contemporary educational psychology, 33(3):360–381. [find](https://scholar.google.com/scholar?q=Noreen+M+Webb%2C+Megan+L+Franke%2C+Marsha+Ing%2C+Angela+Chan%2C+Tondra+De%2C+Deanna+Freund%2C+and+Dan+Battey%2E+2008%2E+The+role+of+teacher+instructional+practices+in+student+collaboration%2E+Contemporary+educational+p)
39. Mikyung Kim Wolf, Amy C Crosson, and Lauren B Resnick. 2005. Classroom talk for rigorous reading comprehension instruction. Reading Psychology, 26(1):27–53. [find](https://scholar.google.com/scholar?q=Mikyung+Kim+Wolf%2C+Amy+C+Crosson%2C+and+Lauren+B+Resnick%2E+2005%2E+Classroom+talk+for+rigorous+reading+comprehension+instruction%2E+Reading+Psychology%2C+26%281%29%3A27%E2%80%9353%2E)
40. Thomas Wolf, Lysandre Debut, Victor Sanh, Julien Chaumond, Clement Delangue, Anthony Moi, Pier-ric Cistac, Tim Rault, Rémi Louf, Morgan Funtowicz, et al. 2019. [Huggingface’s transformers: State-ofthe-art natural language processing](https://md.research.elamur.ai/free/en/papers/wolf-2020-transformers-state-art-natural-language/). arXiv preprint arXiv:1910.03771.
41. Qizhe Xie, Zihang Dai, Eduard Hovy, Thang Luong, and Quoc Le. 2020. [Unsupervised data augmentation for consistency training](https://md.research.elamur.ai/free/en/papers/xie-2019-unsupervised-data-augmentation-consistency-training/). Advances in Neural Information Processing Systems, 33:6256–6268.
42. Zhilin Yang, Zihang Dai, Yiming Yang, Jaime Carbonell, Russ R Salakhutdinov, and Quoc V Le. 2019. [Xlnet: Generalized autoregressive pretraining for language understanding](https://md.research.elamur.ai/free/en/papers/yang-2019-xlnet-generalized-autoregressive-pretraining-language/). Advances in neural information processing systems, 32.

*Updated 3 November 2026*