---
title: "Dynamic Key-Value Memory Networks for Knowledge Tracing"
description: "The authors propose Dynamic Key-Value Memory Networks (DKVMN), a memory-augmented neural… 33 references, cited by 24 papers here."
authors: ["Jiani Zhang", "Xingjian Shi", "Irwin King", "Dit-Yan Yeung"]
year: 2017
venue: "Proceedings of the 26th International Conference on World Wide Web"
source: "https://arxiv.org/abs/1611.08108"
doi: "10.1145/3038912.3052580"
arxiv: "1611.08108"
licence: "arXiv — non-exclusive licence to distribute"
licence_url: "http://arxiv.org/licenses/nonexclusive-distrib/1.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/zhang-2017-dynamic-key-value-memory-networks/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/zhang-2017-dynamic-key-value-memory-networks/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2017](https://md.research.elamur.ai/free/en/papers/2017/) / Dynamic Key-Value Memory Networks for Knowledge Tracing

# Dynamic Key-Value Memory Networks for Knowledge Tracing

**Authors:** [Jiani Zhang](https://md.research.elamur.ai/free/en/authors/jiani-zhang/), [Xingjian Shi](https://md.research.elamur.ai/free/en/authors/xingjian-shi/), [Irwin King](https://md.research.elamur.ai/free/en/authors/irwin-king/) and [Dit-Yan Yeung](https://md.research.elamur.ai/free/en/authors/dit-yan-yeung/) · **Year:** 2017 · **Venue:** [Proceedings of the 26th International Conference on World Wide Web](https://md.research.elamur.ai/free/en/venues/international-world-wide-web-conference-www/) ([WWW 2017](https://md.research.elamur.ai/free/en/venues/international-world-wide-web-conference-www/2017/)) · **Licence:** [arXiv — non-exclusive licence to distribute](http://arxiv.org/licenses/nonexclusive-distrib/1.0/)

**Links:** [Publisher's page](https://doi.org/10.1145/3038912.3052580) · [PDF](https://arxiv.org/pdf/1611.08108) · [DOI 10.1145/3038912.3052580](https://doi.org/10.1145/3038912.3052580) · [arXiv 1611.08108](https://arxiv.org/abs/1611.08108) · [OpenAlex](https://openalex.org/W2559094423)

**Citations:** 862 citations (OpenAlex, 2026-10-14) · [24](https://md.research.elamur.ai/free/en/papers/zhang-2017-dynamic-key-value-memory-networks/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
The paper addresses knowledge tracing (KT): estimating how well a learner understands each underlying concept from their history of attempted exercises and whether each answer was right, then predicting success on the next exercise. Earlier approaches each have a weakness. Bayesian Knowledge Tracing handles every predefined concept in isolation with a binary hidden state. Deep Knowledge Tracing (an LSTM) compresses all knowledge into one hidden vector, which hides per-concept mastery.

The authors propose Dynamic Key-Value Memory Networks (DKVMN), a memory-augmented neural network with two parts:
- a fixed **key** matrix that holds latent concept representations
- a changeable **value** matrix that holds and revises the learner's mastery of each concept.

For each exercise, the model computes correlation weights against the keys, reads only the relevant mastery slots to make its prediction, and then writes updates back to those slots.

Across one synthetic and three real-world datasets, DKVMN consistently beat BKT, DKT and a single-matrix MANN baseline. It also uncovers exercise-to-concept groupings without expert labels and can visualize how a learner's mastery evolves. This makes it useful for adaptive tutoring and MOOCs.

## Key points
- **Architecture:** a static key memory (concepts) paired with a dynamic value memory (per-concept mastery). This differs from standard MANNs, which use either one memory matrix or two static ones.
- **Evaluation:** On four KT datasets, three drawn from real-world sources and one synthetic, DKVMN achieved better results than both BKT and DKT. A plain MANN baseline used cosine-similarity reads and LRUA writes, with the usage decay γ fixed at 0.9.
- **Concept discovery on synthetic-5:** with memory size N = 50, exercises still fell into five concept groups. Assigning each exercise by its maximum correlation weight gave an adjusted mutual information of 0.879, and t-SNE-based clustering reached a perfect 1.0.
- **Qualitative results on real data:** exercises were grouped into 10 clusters. For example, ordering tasks for fractions, integers, decimals and real numbers landed in the same cluster.
- **Knowledge state depiction:** the model can trace individual concept states over time, illustrated with five concepts across 50 answered exercises. Planned future work includes adding exercise content to the embeddings and building hierarchical key-value memories.

*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

- [Hong Kong University of Science and Technology](https://md.research.elamur.ai/free/en/universities/hong-kong-university-of-science-and-technology/) — University, Hong Kong, China · 183 works · 27,600 citations
- [Chinese University of Hong Kong](https://md.research.elamur.ai/free/en/universities/chinese-university-of-hong-kong/) — University, Hong Kong, China · 176 works · 17,536 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 · [🗄️ Information Systems](https://md.research.elamur.ai/free/en/shelves/computer-science/information-systems/) · 1,888 works · [🖥️ Computer Science Applications](https://md.research.elamur.ai/free/en/shelves/computer-science/computer-science-applications/) · 184 works

## Cite

Cite as: Jiani Zhang, Xingjian Shi, Irwin King and Dit-Yan Yeung. 2017. Dynamic Key-Value Memory Networks for Knowledge Tracing. Proceedings of the 26th International Conference on World Wide Web.

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

```bibtex
@inproceedings{zhang2017dynamic,
  title = {{Dynamic Key-Value Memory Networks for Knowledge Tracing}},
  author = {Jiani Zhang, Xingjian Shi, Irwin King and Dit-Yan Yeung},
  booktitle = {Proceedings of the 26th International Conference on World Wide Web},
  year = {2017},
  doi = {10.1145/3038912.3052580},
  eprint = {1611.08108},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.1145/3038912.3052580}
}
```

- [← previous Joint Extraction of Entities and Relations Based on a Novel Tagging Scheme](https://md.research.elamur.ai/free/en/papers/zheng-2017-joint-extraction-entities-relations-based/)
- [next → Challenges in Data-to-Document Generation](https://md.research.elamur.ai/free/en/papers/wiseman-2017-challenges-data-document-generation/)
- [↑ back to papers of 2017](https://md.research.elamur.ai/free/en/papers/2017/)

## References

33 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. A. Bordes, N. Usunier, S. Chopra, and J. Weston. Large-scale simple question answering with memory networks. arXiv preprint arXiv:1506.02075, 2015. [find](https://scholar.google.com/scholar?q=A%2E+Bordes%2C+N%2E+Usunier%2C+S%2E+Chopra%2C+and+J%2E+Weston%2E+Large%2Dscale+simple+question+answering+with+memory+networks%2E+arXiv+preprint+arXiv%3A1506%2E02075%2C+2015%2E)
2. T. Chen, M. Li, Y. Li, M. Lin, N. Wang, M. Wang, T. Xiao, B. Xu, C. Zhang, and Z. Zhang. [Mxnet: A flexible and efficient machine learning library for heterogeneous distributed systems](https://md.research.elamur.ai/free/en/papers/chen-2015-mxnet-flexible-efficient-machine-learning/). In In Neural Information Processing Systems, Workshop on Machine Learning Systems, 2015.
3. A. T. Corbett and J. R. Anderson. Knowledge tracing: Modeling the acquisition of procedural knowledge. volume 4, pages 253–278. Springer, 1994. [find](https://scholar.google.com/scholar?q=A%2E+T%2E+Corbett+and+J%2E+R%2E+Anderson%2E+Knowledge+tracing%3A+Modeling+the+acquisition+of+procedural+knowledge%2E+volume+4%2C+pages+253%E2%80%93278%2E+Springer%2C+1994%2E)
4. R. S. d Baker, A. T. Corbett, and V. Aleven. More accurate student modeling through contextual estimation of slip and guess probabilities in bayesian knowledge tracing. In International Conference on Intelligent Tutoring Systems, pages 406–415. Springer, 2008. [find](https://scholar.google.com/scholar?q=R%2E+S%2E+d+Baker%2C+A%2E+T%2E+Corbett%2C+and+V%2E+Aleven%2E+More+accurate+student+modeling+through+contextual+estimation+of+slip+and+guess+probabilities+in+bayesian+knowledge+tracing%2E+In+International+Conference+on+)
5. M. Feng, N. Heffernan, and K. Koedinger. Addressing the assessment challenge with an online system that tutors as it assesses. User Modeling and User-Adapted Interaction, 19(3):243–266, 2009. [find](https://scholar.google.com/scholar?q=M%2E+Feng%2C+N%2E+Heffernan%2C+and+K%2E+Koedinger%2E+Addressing+the+assessment+challenge+with+an+online+system+that+tutors+as+it+assesses%2E+User+Modeling+and+User%2DAdapted+Interaction%2C+19%283%29%3A243%E2%80%93266%2C+2009%2E)
6. A. Graves, G. Wayne, and I. Danihelka. Neural turing machines. arXiv preprint arXiv:1410.5401, 2014. [find](https://scholar.google.com/scholar?q=A%2E+Graves%2C+G%2E+Wayne%2C+and+I%2E+Danihelka%2E+Neural+turing+machines%2E+arXiv+preprint+arXiv%3A1410%2E5401%2C+2014%2E)
7. A. Graves, G. Wayne, M. Reynolds, T. Harley, I. Danihelka, A. Grabska-Barwińska, S. G. Colmenarejo, E. Grefenstette, T. Ramalho, J. Agapiou, et al. Hybrid computing using a neural network with dynamic external memory. Nature, 538(7626):471–476, 2016. [find](https://scholar.google.com/scholar?q=A%2E+Graves%2C+G%2E+Wayne%2C+M%2E+Reynolds%2C+T%2E+Harley%2C+I%2E+Danihelka%2C+A%2E+Grabska%2DBarwin%CC%81ska%2C+S%2E+G%2E+Colmenarejo%2C+E%2E+Grefenstette%2C+T%2E+Ramalho%2C+J%2E+Agapiou%2C+et+al%2E+Hybrid+computing+using+a+neural+network+with+dynami)
8. E. Grefenstette, K. M. Hermann, M. Suleyman, and P. Blunsom. Learning to transduce with unbounded memory. In Advances in Neural Information Processing Systems, pages 1828–1836, 2015. [find](https://scholar.google.com/scholar?q=E%2E+Grefenstette%2C+K%2E+M%2E+Hermann%2C+M%2E+Suleyman%2C+and+P%2E+Blunsom%2E+Learning+to+transduce+with+unbounded+memory%2E+In+Advances+in+Neural+Information+Processing+Systems%2C+pages+1828%E2%80%931836%2C+2015%2E)
9. S. Hochreiter and J. Schmidhuber. Long short-term memory. Neural computation, 9(8):1735–1780, 1997. [find](https://scholar.google.com/scholar?q=S%2E+Hochreiter+and+J%2E+Schmidhuber%2E+Long+short%2Dterm+memory%2E+Neural+computation%2C+9%288%29%3A1735%E2%80%931780%2C+1997%2E)
10. A. Joulin and T. Mikolov. Inferring algorithmic patterns with stack-augmented recurrent nets. In Advances in Neural Information Processing Systems, pages 190–198, 2015. [find](https://scholar.google.com/scholar?q=A%2E+Joulin+and+T%2E+Mikolov%2E+Inferring+algorithmic+patterns+with+stack%2Daugmented+recurrent+nets%2E+In+Advances+in+Neural+Information+Processing+Systems%2C+pages+190%E2%80%93198%2C+2015%2E)
11. M. Khajah, R. V. Lindsey, and M. C. Mozer. [How deep is knowledge tracing?](https://md.research.elamur.ai/free/en/papers/khajah-2016-deep-knowledge-tracing/) In Educational Data Mining 2016, 2016.
12. K. R. Koedinger, R. S. Baker, K. Cunningham, A. Skogsholm, B. Leber, and J. Stamper. A data repository for the edm community: The pslc datashop. Handbook of educational data mining, 43, 2010. [find](https://scholar.google.com/scholar?q=K%2E+R%2E+Koedinger%2C+R%2E+S%2E+Baker%2C+K%2E+Cunningham%2C+A%2E+Skogsholm%2C+B%2E+Leber%2C+and+J%2E+Stamper%2E+A+data+repository+for+the+edm+community%3A+The+pslc+datashop%2E+Handbook+of+educational+data+mining%2C+43%2C+2010%2E)
13. A. Krizhevsky, I. Sutskever, and G. E. Hinton. [Imagenet classification with deep convolutional neural networks](https://md.research.elamur.ai/free/en/papers/krizhevsky-2012-imagenet-classification-deep-convolutional/). In Advances in Neural Information Processing Systems, pages 1097–1105, 2012.
14. Y. LeCun, Y. Bengio, and G. Hinton. Deep learning. Nature, 521(7553):436–444, 2015. [find](https://scholar.google.com/scholar?q=Y%2E+LeCun%2C+Y%2E+Bengio%2C+and+G%2E+Hinton%2E+Deep+learning%2E+Nature%2C+521%287553%29%3A436%E2%80%93444%2C+2015%2E)
15. L. v. d. Maaten and G. Hinton. Visualizing data using t-sne. Journal of Machine Learning Research, 9(11):2579–2605, 2008. [find](https://scholar.google.com/scholar?q=L%2E+v%2E+d%2E+Maaten+and+G%2E+Hinton%2E+Visualizing+data+using+t%2Dsne%2E+Journal+of+Machine+Learning+Research%2C+9%2811%29%3A2579%E2%80%932605%2C+2008%2E)
16. T. Mikolov, I. Sutskever, K. Chen, G. S. Corrado, and J. Dean. [Distributed representations of words and phrases and their compositionality](https://md.research.elamur.ai/free/en/papers/mikolov-2013-distributed-representations-words-phrases-their/). In Advances in Neural Information Processing Systems, pages 3111–3119, 2013.
17. A. Miller, A. Fisch, J. Dodge, A.-H. Karimi, A. Bordes, and J. Weston. Key-value memory networks for directly reading documents. arXiv preprint arXiv:1606.03126, 2016. [find](https://scholar.google.com/scholar?q=A%2E+Miller%2C+A%2E+Fisch%2C+J%2E+Dodge%2C+A%2E%2DH%2E+Karimi%2C+A%2E+Bordes%2C+and+J%2E+Weston%2E+Key%2Dvalue+memory+networks+for+directly+reading+documents%2E+arXiv+preprint+arXiv%3A1606%2E03126%2C+2016%2E)
18. G. B. Orr and K.-R. Müller. Neural networks: tricks of the trade. Springer, 2003. [find](https://scholar.google.com/scholar?q=G%2E+B%2E+Orr+and+K%2E%2DR%2E+Mu%CC%88ller%2E+Neural+networks%3A+tricks+of+the+trade%2E+Springer%2C+2003%2E)
19. Z. A. Pardos and N. T. Heffernan. Modeling individualization in a bayesian networks implementation of knowledge tracing. In User Modeling, Adaptation, and Personalization, pages 255–266. Springer, 2010. [find](https://scholar.google.com/scholar?q=Z%2E+A%2E+Pardos+and+N%2E+T%2E+Heffernan%2E+Modeling+individualization+in+a+bayesian+networks+implementation+of+knowledge+tracing%2E+In+User+Modeling%2C+Adaptation%2C+and+Personalization%2C+pages+255%E2%80%93266%2E+Springer%2C+201)
20. Z. A. Pardos and N. T. Heffernan. Kt-idem: Introducing item difficulty to the knowledge tracing model. In User Modeling, Adaption and Personalization, pages 243–254. Springer, 2011. [find](https://scholar.google.com/scholar?q=Z%2E+A%2E+Pardos+and+N%2E+T%2E+Heffernan%2E+Kt%2Didem%3A+Introducing+item+difficulty+to+the+knowledge+tracing+model%2E+In+User+Modeling%2C+Adaption+and+Personalization%2C+pages+243%E2%80%93254%2E+Springer%2C+2011%2E)
21. R. Pascanu, T. Mikolov, and Y. Bengio. On the difficulty of training recurrent neural networks. In Proceedings of The 30th International Conference on Machine Learning, pages 1310–1318, 2013. [find](https://scholar.google.com/scholar?q=R%2E+Pascanu%2C+T%2E+Mikolov%2C+and+Y%2E+Bengio%2E+On+the+difficulty+of+training+recurrent+neural+networks%2E+In+Proceedings+of+The+30th+International+Conference+on+Machine+Learning%2C+pages+1310%E2%80%931318%2C+2013%2E)
22. C. Piech, J. Bassen, J. Huang, S. Ganguli, M. Sahami, L. J. Guibas, and J. Sohl-Dickstein. [Deep knowledge tracing](https://md.research.elamur.ai/free/en/papers/piech-2015-deep-knowledge-tracing/). In Advances in Neural Information Processing Systems, pages 505–513, 2015.
23. J. Reye. Student modelling based on belief networks. International Journal of Artificial Intelligence in Education, 14(1):63–96, 2004. [find](https://scholar.google.com/scholar?q=J%2E+Reye%2E+Student+modelling+based+on+belief+networks%2E+International+Journal+of+Artificial+Intelligence+in+Education%2C+14%281%29%3A63%E2%80%9396%2C+2004%2E)
24. A. Santoro, S. Bartunov, M. Botvinick, D. Wierstra, and T. Lillicrap. Meta-learning with memory-augmented neural networks. In Proceedings of The 33rd International Conference on Machine Learning, pages 1842–1850, 2016. [find](https://scholar.google.com/scholar?q=A%2E+Santoro%2C+S%2E+Bartunov%2C+M%2E+Botvinick%2C+D%2E+Wierstra%2C+and+T%2E+Lillicrap%2E+Meta%2Dlearning+with+memory%2Daugmented+neural+networks%2E+In+Proceedings+of+The+33rd+International+Conference+on+Machine+Learning%2C+page)
25. J. Schmidhuber. Deep learning in neural networks: An overview. Neural Networks, 61:85–117, 2015. [find](https://scholar.google.com/scholar?q=J%2E+Schmidhuber%2E+Deep+learning+in+neural+networks%3A+An+overview%2E+Neural+Networks%2C+61%3A85%E2%80%93117%2C+2015%2E)
26. P. Steif and N. Bier. Oli engineering statics - fall 2011. Feb. 2014. [find](https://scholar.google.com/scholar?q=P%2E+Steif+and+N%2E+Bier%2E+Oli+engineering+statics+%2D+fall+2011%2E+Feb%2E+2014%2E)
27. S. Sukhbaatar, J. Weston, R. Fergus, et al. End-to-end memory networks. In Advances in neural information processing systems, pages 2440–2448, 2015. [find](https://scholar.google.com/scholar?q=S%2E+Sukhbaatar%2C+J%2E+Weston%2C+R%2E+Fergus%2C+et+al%2E+End%2Dto%2Dend+memory+networks%2E+In+Advances+in+neural+information+processing+systems%2C+pages+2440%E2%80%932448%2C+2015%2E)
28. O. Vinyals, C. Blundell, T. Lillicrap, D. Wierstra, et al. [Matching networks for one shot learning](https://md.research.elamur.ai/free/en/papers/vinyals-2016-matching-networks-one-shot-learning/). In Advances in Neural Information Processing Systems, pages 3630–3638, 2016.
29. P. Viola and W. M. Wells III. Alignment by maximization of mutual information. International Journal of Computer Vision, 24(2):137–154, 1997. [find](https://scholar.google.com/scholar?q=P%2E+Viola+and+W%2E+M%2E+Wells+III%2E+Alignment+by+maximization+of+mutual+information%2E+International+Journal+of+Computer+Vision%2C+24%282%29%3A137%E2%80%93154%2C+1997%2E)
30. J. Weston, S. Chopra, and A. Bordes. Memory networks. International Conference on Learning Representations, 2015. [find](https://scholar.google.com/scholar?q=J%2E+Weston%2C+S%2E+Chopra%2C+and+A%2E+Bordes%2E+Memory+networks%2E+International+Conference+on+Learning+Representations%2C+2015%2E)
31. K. H. Wilson, X. Xiong, M. Khajah, R. V. Lindsey, S. Zhao, Y. Karklin, E. G. Van Inwegen, B. Han, C. Ekanadham, J. E. Beck, et al. Estimating student proficiency: Deep learning is not the panacea. In In Neural Information Processing Systems, Workshop on Machine Learning for Education, 2016. [find](https://scholar.google.com/scholar?q=K%2E+H%2E+Wilson%2C+X%2E+Xiong%2C+M%2E+Khajah%2C+R%2E+V%2E+Lindsey%2C+S%2E+Zhao%2C+Y%2E+Karklin%2C+E%2E+G%2E+Van+Inwegen%2C+B%2E+Han%2C+C%2E+Ekanadham%2C+J%2E+E%2E+Beck%2C+et+al%2E+Estimating+student+proficiency%3A+Deep+learning+is+not+the+panacea%2E+In+)
32. X. Xiong, S. Zhao, E. G. Van Inwegen, and J. E. Beck. [Going deeper with deep knowledge tracing](https://md.research.elamur.ai/free/en/papers/xiong-2016-going-deeper-deep-knowledge-tracing/). In Educational Data Mining 2016, 2016.
33. M. V. Yudelson, K. R. Koedinger, and G. J. Gordon. Individualized bayesian knowledge tracing models. In Artificial intelligence in education, pages 171–180. Springer, 2013. [find](https://scholar.google.com/scholar?q=M%2E+V%2E+Yudelson%2C+K%2E+R%2E+Koedinger%2C+and+G%2E+J%2E+Gordon%2E+Individualized+bayesian+knowledge+tracing+models%2E+In+Artificial+intelligence+in+education%2C+pages+171%E2%80%93180%2E+Springer%2C+2013%2E)

*Updated 3 November 2026*