Pith. sign in

REVIEW 4 major objections 6 minor 45 references

NR4DER: Neural Re-ranking for Diversified Exercise Recommendation

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read NR4DER claims a three-stage pipeline beats existing exercise recommenders in both accuracy and diversity, reaching NDCG@10 of 0.919 on Nips34.

desk verdict A coherent combination of mLSTM filtering, MELT-style transfer, and RAPID-like re-ranking for exercise recommendation, with large reported gains but an underspecified evaluation and an unverified long-tail transfer mechanism. read the letter →

arxiv 2506.06341 v1 pith:RYIY5TRU submitted 2025-06-01 cs.IR cs.AIcs.CY

Xinghe Cheng , Xufang Zhou , Liangda Fang , Chaobo He , Yuyu Zhou , Weiqi Luo , Zhiguo Gong , Quanlong Guan This is my paper · ORCID
classification cs.IRcs.AIcs.CY
keywords exerciserecommendationneuralre-rankingpersonalizedlearningsequenceaugmentationknowledgetracinglong-tailedstudentdistributionmLSTM
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper is trying to establish that a single exercise-recommendation pipeline can solve two problems at once: the long-tailed student distribution that makes most students' histories too short to model, and the diversity of learning pace that a static recommendation list cannot match. The proposed system, NR4DER, couples an mLSTM-based difficulty filter — whose student-representation enhancer transfers structure learned from active students' recent histories to inactive students — with a neural re-ranker that balances exercise relevance against per-student knowledge-concept diversity. On the Nips34, Assist2009, and Assist2012 datasets the probabilistic variant reaches NDCG@10 of 0.919, 0.775, and 0.816, above the best baseline in each case, and the re-ranking module alone accounts for the largest accuracy gains. A sympathetic reader would care because the long tail is the actual student population in online education: if the transfer and re-ranking claims hold, sparse-history students get better and better-mixed practice lists without any new data from them.

What carries the argument

The engine of the paper is a chain of three modules. The student representation enhancer $G^s_\phi$ is a map trained, on active students only, to reconstruct the full sequence embedding $h_s$ from the embedding $r_s = f(\bar{P}_s)$ of the most recent $T$ interactions, by minimizing $\mathcal{L}_s = w_s |h_s - G^s_\phi(r_s)|^2$ under a sinusoidal weight schedule; it embodies the paper's transfer assumption and is applied to inactive students as $h^+_s = G^s_\phi(r_s) + \beta h_s$. The knowledge concept mastery predictor is an mLSTM — an LSTM whose cell state obeys the covariance update rule $C_t = f_t C_{t-1} + i_t v_t k_t^\top$ — that maps the enhanced representation to mastery per concept $y(k)$; mastery sets exercise difficulty $D_e = 1 - \prod_{k\in e} P(k)$, and a threshold filter keeps the $L$ closest-difficulty exercises as the candidate list. The neural re-ranker, adapted from RAPID, computes a relevance matrix with Bi-LSTM over the candidate list and a diversity vector $\Delta(C_l) = \hat{\omega} \odot d(C_l)$, where $\hat{\omega}$ is the student's per-concept learning-pace distribution from self-attention over per-concept LSTM encodings and $d(C_l)$ is the marginal gain of the submodular coverage function $b_k(C) = 1 - \prod_{e\in C}(1 - \tau^k_e)$; an MLP fuses relevance and diversity into re-ranking scores, either deterministically or through the upper confidence bound $U(C_l) = f^{\varphi}_m(V) + f^s_m(V)$ that the probabilistic variant sorts by.

What would settle it

A concrete test: split inactive students by a behavioral proxy for why their history is short (for example, long inter-session gaps with repeated failures suggest disengagement, while a burst of activity that abruptly stops suggests unfamiliarity) and compare recommendation quality with and without the enhancer within each subgroup — the transfer assumption predicts roughly equal gains, while the alternative predicts gains only in subgroups whose recent activity statistically resembles active students' truncated sequences. A cheaper check: recompute Table 4 with several random seeds; the Assist2012 inactive-student gain (NDCG@10 of 0.811 versus 0.808 with the enhancer) is small enough that error bars may put it inside noise.

Watch

Extended reading notes

Core claim

The paper's central claim is that exercise recommendation should be built as a filter-and-re-rank pipeline that treats sparse student histories and diverse learning paces as first-class problems rather than afterthoughts. In the filter stage, a student representation enhancer is trained on active students alone to reconstruct a complete sequence representation from the most recent $T$ interactions (minimizing $\mathcal{L}_s = w_s |h_s - G^s_\phi(r_s)|^2$), and this reconstruction is blended into each inactive student's own representation as $h^+_s = G^s_\phi(r_s) + \beta h_s$; the enhanced representation feeds an mLSTM-based knowledge-concept mastery predictor, whose outputs set exercise difficulty and drive a difficulty-threshold filter. In the re-ranking stage, a Bi-LSTM relevance estimator scores each candidate exercise in list context while a learning-pattern diversity estimator multiplies a per-knowledge-concept pace distribution by the marginal gain of a submodular coverage function, and an MLP fuses the two, with a probabilistic variant that ranks by upper confidence bound. The full system outperforms all seven baselines on NDCG across all three datasets and on most other metrics too, with the probabilistic variant's NDCG@10 reaching 0.919 on Nips34, 0.775 on Assist2009, and 0.816 on Assist2012; the paper concedes one exception, lower F1 on Assist2009, which it attributes to F1 rewarding classification accuracy over ranking quality.

Load-bearing premise

The load-bearing premise is that an inactive student's short practice history looks like the most recent stretch of an active student's history, so a network trained to reconstruct full representations from truncated active sequences can safely enrich inactive students' representations. If students are inactive for qualitatively different reasons — disengagement, unfamiliarity with the platform, or a different starting level — the reconstruction can distort rather than improve their representations, and the reported benefit for inactive students would not transfer to those groups.

Editorial extensions

If this is right

  • If the central claim holds, a single pipeline can raise both ranking accuracy (NDCG, Recall) and list diversity for practice exercises, weakening the usual accuracy–diversity tradeoff in this domain.
  • The active-to-inactive transfer (Table 4) implies that sparse-history students can be served better without waiting for them to accumulate more data — the recent-history structure of active students carries usable signal for the long tail.
  • The per-knowledge-concept pace distribution ($\hat{\omega}$ in Eq. 12) means the system can produce different mixes for different students: concept-diverse lists for broad learners, focused reinforcement lists for students consolidating specific skills, instead of one static diverse list.
  • Concretely, the probabilistic variant NR4DER-p reaches NDCG@10 of 0.919 on Nips34, 0.775 on Assist2009, and 0.816 on Assist2012, against best-baseline values of 0.846, 0.732, and 0.603.
  • The re-ranking module, not just better mastery prediction, drives most of the accuracy gain: adding it raises NDCG@10 from 0.613 to 0.919 on Nips34 (Table 5), suggesting list-context scoring is where the headroom is.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The transfer trick — reconstruct a full-user representation from a truncated recent window — is not specific to exercise data; it could be borrowed by any sequential recommender with a long-tailed user base, but the paper validates it only on the three education datasets.
  • The pace vector $\hat{\omega} \odot d(C_l)$ is, in effect, a per-user submodular diversification signal; a teacher-facing tool that surfaces it as an interpretable learning-pace profile would be a direct, untested application.
  • The probabilistic re-ranker's upper confidence bound behaves as exploration, so an untested extension is to scale exploration with progress — recommend a wider mix while a student is stagnating, a tighter mix while they improve.
  • Because evaluation is on historical logs with proxy metrics (NDCG, Recall, F1, DIV), the paper does not establish learning-outcome gains; showing that such lists reduce dropout or improve retention would require a live or counterfactual intervention study.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. NR4DER is a three-stage pipeline for exercise recommendation. A student representation enhancer, based on MELT [17], trains a reconstruction network G_phi on truncated sequences of active students and applies it to inactive students to improve their representations. An mLSTM knowledge-concept mastery predictor computes exercise difficulty, and a filter retains the L exercises closest to a difficulty threshold. A neural re-ranker, adapted from RAPID [20], combines Bi-LSTM relevance scores with a learning-pace diversity gain to produce the final top-k list. Experiments on Nips34, Assist2009, and Assist2012 compare with seven baselines and report large gains (e.g., NDCG@10 of 0.919 versus 0.846 for the best baseline on Nips34), plus ablations for the enhancer and the re-ranker. The central claims are that NR4DER significantly improves both accuracy and diversity and alleviates the long-tailed student distribution problem.

Significance. The paper ships reproducible code and applies two proven mechanisms (MELT's long-tail enhancement and RAPID's re-ranking) in a new domain, which is a plausible route to improving both accuracy and diversity. The reported gains are large and the ablation structure is sensible. Nevertheless, the current evidence is conditional: the evaluation protocol is underspecified, the DIV metric is undefined, no uncertainty or significance is reported, and the long-tail mechanism shows only small inactive-user improvements in its own ablation. If the authors can close these gaps, the result would be a solid systems-oriented contribution to educational recommendation.

major comments (4)
  1. [Section 5.3; Tables 3-5] The evaluation protocol is not specified enough to support the abstract's 'significantly outperforms' claim. The paper never defines the ground-truth relevance labels for NDCG/Recall/F1 (next exercise? held-out unseen exercises? exercises not yet mastered?), the negative-sampling or candidate-generation procedure for the filter and re-ranker, or the formula for the DIV metric used in Figure 3. Eq. (19) additionally requires binary labels for 'mastered' versus 'not mastered' exercises, but no source for these labels is given, and the values of L and of the final hyperparameters are not reported. Please specify the full protocol, report means and standard deviations over multiple runs, and include significance tests; otherwise the magnitude of the reported gains cannot be assessed.
  2. [Section 4.1.1; Table 4] The load-bearing assumption that the most recent T interactions of an active student are representative of an inactive student's full sequence is never tested, and T is not reported. Table 4 shows inactive-only NDCG@10 gains of +0.022 (Nips34), +0.034 (Assist2009), and +0.003 (Assist2012), with the Assist2012 gain essentially zero, while active-only gains are comparable or larger. This does not demonstrate that the enhancer solves the long-tail problem. At inference, r_s = f(P_s) for an inactive student is computed from a sequence whose length may be far below T, so the input distribution at inference need not match the training distribution. Please report active/inactive gains with uncertainty, state T and the length distribution of inactive sequences, and test the transfer assumption directly, for example by checking whether G_phi(r_s) improves reconstruction of held-out inactive representations compared to using r_s alone.
  3. [Section 4.2; Table 5] The w/o NR ablation is difficult to interpret. On Nips34, removing the neural re-ranker reduces NDCG@10 from 0.919 to 0.613 while F1@10 stays nearly unchanged (0.727 versus 0.726); similar patterns appear on the other datasets. Please clarify what the w/o NR system actually outputs (e.g., ranking by predicted difficulty only), explain the large NDCG/F1 divergence, and report the DIV metric for both variants so that the diversity claim in RQ3 is directly supported.
  4. [Section 3.2.2; Eq. (2)] The difficulty definition is ill-defined as written. P(k) is introduced as the vector [P(k1),...,P(km)], but Eq. (2) takes a product over k in e of P(k), mixing a vector with scalar quantities. Since Eq. (11) and the exercise filter use D_e to build the candidate set C, this definition must be corrected and stated unambiguously, for example as D_e = 1 - product_{k in e} p_k with scalar per-concept mastery probabilities p_k.
minor comments (6)
  1. [Section 4.1.1; Eq. (5), Eq. (7)] The enhancer is written G^s_phi in Eq. (5) but G_phi in Eq. (7); please use one symbol consistently.
  2. [Section 5.4; Eq. (10)] Section 5.4 uses gamma_s for the coefficient that Eq. (10) calls lambda_s; please reconcile the notation.
  3. [Table 3] The Assist2012 rows abbreviate DKTRec/AKTRec as DKT/AKT, while the other datasets use the full names; this should be made consistent.
  4. [Figures 3 and 4] The axis labels in Figures 3 and 4 render as garbled escaped character sequences in the submitted manuscript, making the diversity results and the knowledge-concept visualization unreadable; please provide clean figures.
  5. [Section 3.2.1] The statement that mLSTM 'performs better in sequential tasks compared to RNN-based methods and Transformer-based methods' is given without a citation or controlled comparison; please either cite the relevant benchmark or soften the claim.
  6. [Section 4.1.1; Eq. (6)] The loss coefficient w_s can become zero at the final epoch for the longest active students because the sine argument reaches pi; please confirm this is intended and describe the schedule's behavior.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the evaluation is external, and the transfer assumption is a correctness concern rather than a definitional or fitted-input reduction.

full rationale

The paper's central claim, that NR4DER outperforms existing methods, is tested on three external datasets (Nips34, Assist2009, Assist2012) against seven baselines in Table 3, with no parameter fitted to the headline metrics in a way that forces the outcome. The two main modules are explicitly adopted from external prior work: the student representation enhancer is attributed to MELT [17] (Section 4.1.1, Eqs. 3-7), and the neural re-ranking module is motivated by RAPID [20] (Section 4.2). Neither module is justified by a self-citation, and the paper's own prior work (KG4Ex [12], KG4EER [11]) appears only as related work and as a baseline, so it is not load-bearing. The enhancer's assumption that a truncated active sequence resembles an inactive student's sequence is an unverified distributional premise, as the skeptical note observes, but this is a correctness and robustness risk rather than a circular step: the training objective in Eq. 5 is not defined in terms of the evaluation metrics, and the reported gains are empirical rather than entailed by construction. Similarly, using the diversity metric from [20] is a standard external evaluation choice and does not make the benchmark circular. No step reduces to its own inputs under the quoted-reduction standard, so the appropriate finding is no significant circularity.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central claim rests on a standard ML training pipeline plus three domain assumptions: concept independence in difficulty calculation, transferability of active-student subsequences to inactive students, and the use of not-mastered labels as the training signal. The free parameters are mostly hyperparameters whose final values are not reported, plus the difficulty threshold inherited from prior work. No new physical or conceptual entities are introduced.

free parameters (6)
  • beta = searched over {0.4,0.6,0.8,1.0}; final value not reported
    Blends original and enhanced student representations in Eq. (7).
  • lambda_s (also written gamma_s) = searched over {0.1,0.3,0.5,0.7,1.0}; final value not reported
    Weights the representation-enhancement loss in Eq. (10); notation is inconsistent between Section 4.1.3 and Section 5.4.
  • delta = 0.7
    Difficulty threshold for filtering candidate exercises in Eq. (11), taken from KCPER [37].
  • T = not reported
    Number of most recent interactions used to truncate active students' sequences in Eq. (3); a central design choice never stated.
  • L = not reported
    Size of the candidate exercise subset selected by the filter; defined in Section 4.1.4 but no value given in experiments.
  • learning rate, batch size, attention heads = lr=0.001, batch in {16,32,64,128}, heads in {2,4,6,8}; final selections not reported
    Standard hyperparameters tuned before final runs (Section 5.4).
assumptions (5)
  • domain assumption Independence of knowledge concepts in difficulty estimation
    Eq. (2) computes exercise difficulty as 1 - product of concept mastery probabilities, treating correctness on each concept as independent; if concepts interact, difficulty estimates are biased.
  • domain assumption Transferability of active-student recent subsequences to inactive students
    Section 4.1.1 assumes inactive students' short sequences are similar to the most recent T interactions of active students, so a mapping trained on active students can enhance inactive representations.
  • domain assumption Not-mastered exercises are the appropriate recommendation targets
    The re-ranker loss (Eq. 19) labels exercises as positive if they are not mastered; this equates recommendation quality with targeting non-mastered items and ignores other signals such as engagement or actual next-exercise choices.
  • standard math Standard sequence-modeling machinery (mLSTM, Bi-LSTM, attention) can represent learning patterns
    The method relies on learned representations from standard architectures without assumptions beyond their usual inductive biases.
  • domain assumption The 8:2 train/test split yields a valid evaluation
    Section 5.4 reports an 8:2 split but does not describe a separate validation set or temporal split; if interactions overlap in time, future exercises may be predicted using later information (leakage).

how reviews work

0 comments
Cite this review

Pith. "Pith review of NR4DER: Neural Re-ranking for Diversified Exercise Recommendation." pith.science (2026). https://pith.science/paper/RYIY5TRU

@misc{pith2026250606341,
  author       = {Pith},
  title        = {Pith review of: NR4DER: Neural Re-ranking for Diversified Exercise Recommendation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RYIY5TRU}},
  note         = {Machine review of arXiv:2506.06341}
}
read the original abstract

With the widespread adoption of online education platforms, an increasing number of students are gaining new knowledge through Massive Open Online Courses (MOOCs). Exercise recommendation have made strides toward improving student learning outcomes. However, existing methods not only struggle with high dropout rates but also fail to match the diverse learning pace of students. They frequently face difficulties in adjusting to inactive students' learning patterns and in accommodating individualized learning paces, resulting in limited accuracy and diversity in recommendations. To tackle these challenges, we propose Neural Re-ranking for Diversified Exercise Recommendation (in short, NR4DER). NR4DER first leverages the mLSTM model to improve the effectiveness of the exercise filter module. It then employs a sequence enhancement method to enhance the representation of inactive students, accurately matches students with exercises of appropriate difficulty. Finally, it utilizes neural re-ranking to generate diverse recommendation lists based on individual students' learning histories. Extensive experimental results indicate that NR4DER significantly outperforms existing methods across multiple real-world datasets and effectively caters to the diverse learning pace of students.

Figures

Figures reproduced from arXiv: 2506.06341 by the authors.

Figure 1
Figure 1. (a) Distribution of interaction frequencies and diver [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. NR4DER Framework. The exercise filter module (a) takes students’ historical exercise sequence as input and generates [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The diversity of three datasets. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 (a) Behavior history Sphere Vol. Int. Mult/Div Prob. Concept Proportions Int. Add 2 Events Prob. Lin. Equations Int. Sub Abs. Value Unit Rate Fraction Div. Percentages Lin. Eq. Systems Mult. Fractions Mode Value Single Event Prob. Trapezoid Area Alg. Solving Rate Value Square Root 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 (b… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: The distribution of knowledge concepts. 5.7 (RQ3) Diversity Analysis To comprehensively evaluate the diversity of NR4DER in gener￾ating exercise recommendations, we introduce a diversity metric. Notably, KT-based methods and MMER models generate static rec￾ommendation …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

45 extracted references · 42 canonical work pages

  1. [17]

    Kibum Kim, Dongmin Hyun, Sukwon Yun, and Chanyoung Park. 2023. MELT: Mutual Enhancement of Long-Tailed User and Item for Sequential Recommenda- tion. InProceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR-2023). 68–77

    work page 2023

  2. [20]

    Weiwen Liu, Yunjia Xi, Jiarui Qin, Xinyi Dai, Ruiming Tang, Shuai Li, Weinan Zhang, and Rui Zhang. 2023. Personalized Diversification for Neural Re-ranking in Recommendation. InProceedings of the 39th IEEE International Conference on Data Engineering (ICDE-2023). 802–815

    work page 2023

  3. [1]

    Maximilian Beck, Korbinian Pöppel, Markus Spanring, Andreas Auer, Oleksandra Prudnikova, Michael Kopp, Günter Klambauer, Johannes Brandstetter, and Sepp Hochreiter. 2024. xLSTM: Extended Long Short-Term Memory.arXiv preprint arXiv:2405.04517(2024). SIGIR ’25, July 13–18, 2025, Padua, Italy. Xinghe Cheng et al

    arXiv 2024

  4. [2]

    Inma Borrella, Sergio Caballero-Caballero, and Eva Ponce-Cueto. 2022. Tak- ing Action to Reduce Dropout in MOOCs: Tested Interventions.Computers & Education179 (2022), 104412

    work page 2022

  5. [3]

    Nancy Burruss and A Popkess. 2013. The Diverse Learning Needs of Students. Teaching in Nursing: A Guide for Faculty(2013), 15–33

    work page 2013

  6. [4]

    Jaime Carbonell and Jade Goldstein. 1998. The Use of MMR, Diversity-Based Reranking for Reordering Documents and Producing Summaries. InProceedings of the 21st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR-1998). 335–336

    work page 1998

  7. [5]

    Xiaojing Fan, Chunliang Tao, and Jianyu Zhao. 2024. Advanced Stock Price Prediction with Xlstm-based Models: Improving Long-term Forecasting.Preprints 2024082109 (2024)

    work page 2024

  8. [6]

    Ziwei Fan, Zhiwei Liu, Shen Wang, Lei Zheng, and Philip S Yu. 2021. Modeling Sequences as Distributions with Uncertainty for Sequential Recommendation. InProceedings of the 30th ACM International Conference on Information and Knowledge Management (CIKM-2021). 3019–3023

    work page 2021

Show all 45 references
  1. [7]

    Mingyu Feng, Neil Heffernan, and Kenneth Koedinger. 2009. Addressing the Assessment Challenge With an Online System That Tutors as It Assesses.User Modeling and User-adapted Interaction19 (2009), 243–266

    2009

  2. [8]

    Wenzheng Feng, Jie Tang, and Tracy Xiao Liu. 2019. Understanding Dropouts in MOOCs. InProceedings of the 33rd AAAI Conference on Artificial Intelligence (AAAI-2019). 517–524

    2019

  3. [9]

    Yue Feng, Jun Xu, Yanyan Lan, Jiafeng Guo, Wei Zeng, and Xueqi Cheng. 2018. From Greedy Selection to Exploratory Decision-Making: Diverse Ranking with Policy-Value Networks. InProceedings of the 41st International ACM SIGIR Confer- ence on Research & Development in Information...

    2018

  4. [10]

    Aritra Ghosh, Neil Heffernan, and Andrew S Lan. 2020. Context-Aware Atten- tive Knowledge Tracing. InProceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (SIGKDD-2020). 2330–2339

    2020

  5. [11]

    Quanlong Guan, Xinghe Cheng, Fang Xiao, Zhuzhou Li, Chaobo He, Liangda Fang, Guanliang Chen, Zhiguo Gong, and Weiqi Luo. 2025. Explainable Exercise Recommendation with Knowledge Graph.Neural Networks183 (2025), 106954

    2025

  6. [12]

    Quanlong Guan, Fang Xiao, Xinghe Cheng, Liangda Fang, Ziliang Chen, Guan- liang Chen, and Weiqi Luo. 2023. Kg4ex: An Explainable Knowledge Graph-Based Approach for Exercise Recommendation. InProceedings of the 32nd ACM Inter- national Conference on Information and Knowledge Ma...

    2023

  7. [13]

    S Hochreiter. 1997. Long Short-term Memory.Neural Computation9, 8 (1997), 1735–1780

    1997

  8. [14]

    Zhenya Huang, Qi Liu, Chengxiang Zhai, Yu Yin, Enhong Chen, Weibo Gao, and Guoping Hu. 2019. Exploring Multi-Objective Exercise Recommendations in Online Education Systems. InProceedings of the 28th ACM International Conference on Information and Knowledge Management (CIKM-201...

    2019

  9. [15]

    Yujia Huo, Derek F Wong, Lionel M Ni, Lidia S Chao, and Jing Zhang. 2020. Knowl- edge Modeling via Contextualized Representations for LSTM-based Personalized Exercise Recommendation.Information Sciences523 (2020), 266–278

    2020

  10. [16]

    Seongwon Jang, Hoyeop Lee, Hyunsouk Cho, and Sehee Chung. 2020. Cities: Contextual Inference of Tail-Item Embeddings for Sequential Recommendation. InProceedings of the 20th IEEE International Conference on Data Mining (ICDM- 2020). 202–211

    2020

  11. [18]

    Fei Liu, Xuegang Hu, Shuochen Liu, Chenyang Bu, and Le Wu. 2023. Meta Multi-Agent Exercise Recommendation: A Game Application Perspective. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (SIGKDD-2023). 1441–1452

    2023

  12. [19]

    Siyi Liu and Yujia Zheng. 2020. Long-Tail Session-Based Recommendation. In Proceedings of the 14th ACM Conference on Recommender Systems (RecSys-2020). 509–514

    2020

  13. [21]

    Zhiwei Liu, Ziwei Fan, Yu Wang, and Philip S Yu. 2021. Augmenting Sequential Recommendation with Pseudo-Prior Items via Reversely Pre-Training Trans- former. InProceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGI...

    2021

  14. [22]

    Zitao Liu, Qiongqiong Liu, Jiahao Chen, Shuyan Huang, and Weiqi Luo. 2023. SimpleKT: a Simple but Tough-to-Beat Baseline for Knowledge Tracing. InThe Eleventh International Conference on Learning Representations (ICLR-2023)

    2023

  15. [23]

    Ilya Loshchilov and Frank Hutter. 2016. Sgdr: Stochastic Gradient Descent with Warm Restarts.arXiv preprint arXiv:1608.03983(2016)

    2016 arXiv

  16. [24]

    Otilia Maria Alejandro Molina, Gabriel Limones Obando, Ashley Aviles Bastidas, and Elias Madrid Mosquera. 2022. Motivation: Personalization of Contents to Reduce Dropout Rates. In2022 IEEE Learning with MOOCS. IEEE, 61–66

    2022

  17. [25]

    Liang Pang, Jun Xu, Qingyao Ai, Yanyan Lan, Xueqi Cheng, and Jirong Wen

  18. [26]

    Chris Piech, Jonathan Bassen, Jonathan Huang, Surya Ganguli, Mehran Sahami, Leonidas J Guibas, and Jascha Sohl-Dickstein. 2015. Deep Knowledge Tracing. In Advances in Neural Information Processing Systems 28 (NeurIPS-2015). 505–513

    2015

  19. [27]

    Yimeng Ren, Kun Liang, Yuhu Shang, and Yiying Zhang. 2023. MulOER-SAN: 2-layer Multi-objective Framework for Exercise Recommendation with Self- attention Networks.Knowledge-Based Systems260 (2023), 110117

    2023

  20. [28]

    Mike Schuster and Kuldip K Paliwal. 1997. Bidirectional Recurrent Neural Net- works.IEEE transactions on Signal Processing45, 11 (1997), 2673–2681

    1997

  21. [29]

    Wenzhuo Song, Shoujin Wang, Yan Wang, and Shengsheng Wang. 2021. Next- item Recommendations in Short Sessions. InProceedings of the 15th ACM Confer- ence on Recommender Systems (RecSys-2021). 282–291

    2021

  22. [30]

    María Cora Urdaneta-Ponte, Amaia Mendez-Zorrilla, and Ibon Oleagordia-Ruiz

  23. [31]

    A Vaswani. 2017. Attention Is All You Need. InAdvances in Neural Information Processing Systems 30 (NeurIPS-2017). 5998–6008

    2017

  24. [32]

    Fan Wang, Xiaomin Fang, Lihang Liu, Yaxue Chen, Jiucheng Tao, Zhiming Peng, Cihang Jin, and Hao Tian. 2019. Sequential Evaluation and Generation Framework for Combinatorial Recommender System.arXiv preprint arXiv:1902.00245(2019)

    2019 arXiv

  25. [33]

    Jiapu Wang, Zheng Cui, Boyue Wang, Shirui Pan, Junbin Gao, Baocai Yin, and Wen Gao. 2024. IME: Integrating Multi-curvature Shared and Specific Embedding for Temporal Knowledge Graph Completion. InProceedings of the ACM Web Conference 2024 (WWW-2024). 1954–1962

    2024

  26. [34]

    Jiapu Wang, Sun Kai, Linhao Luo, Wei Wei, Yongli Hu, Alan Wee-Chung Liew, Shirui Pan, and Baocai Yin. 2024. Large Language Models-guided Dynamic Adaptation for Temporal Knowledge Graph Reasoning. InAdvances in Neural Information Processing Systems 38 (NeurIPS-2024). 8384–8410

    2024

  27. [35]

    Jiapu Wang, Boyue Wang, Junbin Gao, Shirui Pan, Tengfei Liu, Baocai Yin, and Wen Gao. 2024. Made: Multicurvature Adaptive Embedding for Temporal Knowl- edge Graph Completion.IEEE Transactions on Cybernetics(2024)

    2024

  28. [36]

    Billy Tak Ming Wong and Kam Cheong Li. 2020. Meeting Diverse Student Needs for Support Services: A Comparison Between Face-to-Face and Distance-Learning Students.Innovating Education in Technology-Supported Environments(2020), 253–268

    2020

  29. [37]

    Zhengyang Wu, Ming Li, Yong Tang, and Qingyu Liang. 2020. Exercise Recom- mendation Based on Knowledge Concept Prediction.Knowledge-Based Systems 210 (2020), 106481

    2020

  30. [38]

    Jianwen Yin, Chenghao Liu, Weiqing Wang, Jianling Sun, and Steven CH Hoi

  31. [39]

    Yisong Yue and Carlos Guestrin. 2011. Linear Submodular Bandits and Their Application to Diversified Retrieval. InAdvances in Neural Information Processing Systems 24 (NeurIPS-2011). 2483–2491

    2011

  32. [40]

    Sukwon Yun, Kibum Kim, Kanghoon Yoon, and Chanyoung Park. 2022. Lte4g: Long-Tail Experts for Graph Neural Networks. InProceedings of the 31st ACM International Conference on Information and Knowledge Management (CIKM-2022). 2434–2443

    2022

  33. [41]

    InProceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (SIGKDD-2020)

    Learning Transferrable Parameters for Long-Tailed Sequential User Behav- ior Modeling. InProceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (SIGKDD-2020). 359–367

    2020

  34. [42]

    Tao Zhuang, Wenwu Ou, and Zhirong Wang. 2018. Globally Optimized Mutual Influence Aware Ranking in E-Commerce Search. InProceedings of the 27th International Joint Conference on Artificial Intelligence (IJCAI-2018). 3725–3731

    2018

  35. [44]

    Jiani Zhang, Xingjian Shi, Irwin King, and Dit-Yan Yeung. 2017. Dynamic Key- Value Memory Networks for Knowledge Tracing. InProceedings of the 26th International Conference on World Wide Web (WWW-2017). 765–774

    2017

  36. [2020]

    InProceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR-2020)

    Setrank: Learning a Permutation-Invariant Ranking Model for Informa- tion Retrieval. InProceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR-2020). 499–508

    2020

  37. [2021]

    Recommendation Systems for Education: Systematic Review.Electronics 10, 14 (2021), 1611

    2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.