pith. sign in

arxiv: 2602.17036 · v3 · submitted 2026-02-19 · 💻 cs.IR · cs.LG

LiveGraph: Active-Structure Neural Re-ranking for Exercise Recommendation

Rong Fu , Zijian Zhang , Haiyun Wei , Jiekai Wu , Kun Liu , Xianda Li , Haoyu Zhao , Yang Li
show 4 more authors
Yongtai Liu Ziming Wang Rui Lu Simon Fong
This is my paper

Pith reviewed 2026-05-15 21:22 UTC · model grok-4.3

classification 💻 cs.IR cs.LG
keywords exercise recommendationneural re-rankinggraph representation learningeducational data miningrecommendation diversitypersonalized learningstudent engagement
0
0 comments X

The pith

LiveGraph uses graph structures to recommend exercises that are both accurate and diverse for students at different engagement levels.

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

The paper presents LiveGraph as a neural re-ranking system for exercise recommendations in digital learning settings. It builds enhanced representations of student activity by linking learning histories into graphs, which helps connect data from highly active students to those who engage less often. A dynamic re-ranking step then promotes a wider range of exercises instead of similar repeats. Sympathetic readers would care because such a system could support more balanced learning progress by addressing uneven participation without needing extra data beyond typical platform logs.

Core claim

LiveGraph is an active-structure neural re-ranking framework that utilizes a graph-based representation enhancement strategy to bridge the information gap between active and inactive students while integrating a dynamic re-ranking mechanism to foster content diversity, thereby balancing recommendation precision with pedagogical variety.

What carries the argument

LiveGraph, an active-structure neural re-ranking framework that builds graph-enhanced student representations from learning histories and applies dynamic re-ranking to increase exercise variety.

If this is right

  • Surpasses contemporary baselines in predictive accuracy across multiple real-world datasets.
  • Increases the breadth of exercise diversity delivered in recommendations.
  • Addresses the long-tailed distribution of student engagement patterns.
  • Adapts recommendations to individual learning trajectories while maintaining precision.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The graph approach could extend to other recommendation settings with similar imbalances in user activity, such as content suggestions for occasional users.
  • Further tests on datasets with very sparse connections among inactive students would clarify practical limits.
  • Pairing the re-ranking with explicit pedagogical goals like skill progression could strengthen outcomes beyond diversity counts.

Load-bearing premise

The graph-based representation enhancement strategy bridges the information gap between active and inactive students without introducing new biases or requiring data beyond what is typically available in learning platforms.

What would settle it

Running the model without the graph enhancement component and finding no accuracy gain specifically for inactive students on the same datasets would show the central claim does not hold.

Figures

Figures reproduced from arXiv: 2602.17036 by Haiyun Wei, Haoyu Zhao, Jiekai Wu, Kun Liu, Rong Fu, Rui Lu, Simon Fong, Xianda Li, Yang Li, Yongtai Liu, Zijian Zhang, Ziming Wang.

Figure 1
Figure 1. Figure 1: Overview of the LiveGraph framework for active-structure neural exercise recommendation. The Graph￾Aware Student Representation Enhancer (Graph-VAE) transforms interaction history Ds into a stochastic mastery distribution θs, regularized by LLM-informed priors. In the Uncertainty-Aware Neural Re-ranker, candidate exercises C are scored based on a multi-signal fusion of relevance ϕrel, diversity ϕdiv, and B… view at source ↗
Figure 2
Figure 2. Figure 2: Convergence behaviour: training steps vs. validation NDCG@5. Curves compare runs [PITH_FULL_IMAGE:figures/full_fig_p010_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Diversity comparison across methods. Grouped bars show DIV@1, DIV@3, DIV@5 and DIV@10 for [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Knowledge–concept distribution for a sampled student (Assist2009). The horizontal axis lists the 20 [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Hyperparameter sensitivity of the VAE prior coefficient [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
read the original abstract

The continuous expansion of digital learning environments has catalyzed the demand for intelligent systems capable of providing personalized educational content. While current exercise recommendation frameworks have made significant strides, they frequently encounter obstacles regarding the long-tailed distribution of student engagement and the failure to adapt to idiosyncratic learning trajectories. We present LiveGraph, a novel active-structure neural re-ranking framework designed to overcome these limitations. Our approach utilizes a graph-based representation enhancement strategy to bridge the information gap between active and inactive students while integrating a dynamic re-ranking mechanism to foster content diversity. By prioritizing the structural relationships within learning histories, the proposed model effectively balances recommendation precision with pedagogical variety. Comprehensive experimental evaluations conducted on multiple real-world datasets demonstrate that LiveGraph surpasses contemporary baselines in both predictive accuracy and the breadth of exercise diversity.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript introduces LiveGraph, an active-structure neural re-ranking framework for exercise recommendation in digital learning platforms. It proposes a graph-based representation enhancement to address long-tailed student engagement and information gaps between active and inactive users, combined with a dynamic re-ranking mechanism to improve content diversity, and reports superior predictive accuracy and exercise variety over contemporary baselines across multiple real-world datasets.

Significance. If the experimental claims are substantiated with full methodological details, the work could offer a practical contribution to educational recommender systems by balancing precision with pedagogical diversity in graph-structured interaction data. However, the absence of architecture specifications, training protocols, and statistical validation in the current presentation limits evaluation of whether the reported gains exceed standard graph neural recommendation practices.

major comments (1)
  1. [Abstract and Experimental Evaluation] Abstract and Experimental Evaluation section: the claim that LiveGraph 'surpasses contemporary baselines in both predictive accuracy and the breadth of exercise diversity' is presented without any description of model architecture, embedding dimensions, training procedure, data splits, or statistical significance testing, preventing verification of the central empirical result.
minor comments (2)
  1. [Method] Clarify the precise definition of 'active-structure' and how the dynamic re-ranking mechanism differs from standard graph attention or message-passing layers used in prior recommendation literature.
  2. [Experimental Setup] Ensure all dataset statistics (number of students, exercises, interactions, sparsity) and baseline implementations are fully reported to allow reproduction.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. The comment correctly identifies that the abstract's brevity limits immediate verification of empirical claims, and we will revise to ensure full methodological transparency while preserving the paper's core contributions.

read point-by-point responses

  1. Referee: [Abstract and Experimental Evaluation] Abstract and Experimental Evaluation section: the claim that LiveGraph 'surpasses contemporary baselines in both predictive accuracy and the breadth of exercise diversity' is presented without any description of model architecture, embedding dimensions, training procedure, data splits, or statistical significance testing, preventing verification of the central empirical result.

    Authors: We agree that the abstract, constrained by length, omits these specifics and that the experimental section requires clearer exposition for independent verification. In the revised version, we will expand the Experimental Evaluation section with a dedicated implementation subsection detailing the graph neural network architecture (active-structure enhancement layers), embedding dimensions, training procedure (including optimizer, learning rates, and epochs), data split protocol, and statistical significance testing (e.g., paired tests with p-values). We will also revise the abstract to briefly reference key methodological elements. These changes will directly address the verification concern without altering the reported results. revision: yes

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The LiveGraph paper describes a graph-enhanced neural re-ranker for exercise recommendation using standard nodes (students/exercises) and interaction edges, with dynamic re-ranking for diversity. No equations, self-definitional steps, or load-bearing self-citations are present in the provided abstract or description that would reduce any claimed prediction or result to its own inputs by construction. Experimental superiority is asserted via evaluations on real-world datasets following established recommendation-system patterns, with no evidence of fitted parameters being relabeled as independent predictions or uniqueness theorems imported from prior author work. The derivation chain remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The approach rests on standard domain assumptions from graph-based recommendation systems with no new entities postulated and multiple free parameters typical of neural models.

free parameters (1)
  • neural network hyperparameters and embedding dimensions
    Fitted during training on student interaction data to optimize the graph representation and re-ranking scores.
axioms (1)
  • domain assumption Graph structures derived from learning histories can transfer useful information between active and inactive students
    Central to the representation enhancement strategy described in the abstract.

pith-pipeline@v0.9.0 · 5454 in / 1269 out tokens · 33509 ms · 2026-05-15T21:22:56.918847+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

73 extracted references · 73 canonical work pages

  1. [1]

    Knowpath: An llm- supported knowledge graph construction and path finding framework to explainable mooc recommendations

    Jia Zhu, Zhangze Chen, Pasquale De Meo, Jueqi Guan, Zhongmei Han, and Weijie Shi. Knowpath: An llm- supported knowledge graph construction and path finding framework to explainable mooc recommendations. ACM Transactions on Information Systems, 44(2):1–28, 2026

    work page 2026

  2. [2]

    Education-oriented graph retrieval-augmented generation for learning path recommendation.arXiv preprint arXiv:2506.22303, 2025

    Xinghe Cheng, Zihan Zhang, Jiapu Wang, Liangda Fang, Chaobo He, Quanlong Guan, Shirui Pan, and Weiqi Luo. Education-oriented graph retrieval-augmented generation for learning path recommendation.arXiv preprint arXiv:2506.22303, 2025

    work page arXiv 2025

  3. [3]

    Towards long-tailed recognition for graph classification via collaborative experts.IEEE Transactions on Big Data, 9(6):1683–1696, 2023

    Si-Yu Yi, Zhengyang Mao, Wei Ju, Yong-Dao Zhou, Luchen Liu, Xiao Luo, and Ming Zhang. Towards long-tailed recognition for graph classification via collaborative experts.IEEE Transactions on Big Data, 9(6):1683–1696, 2023

    work page 2023

  4. [4]

    Lte4g: Long-tail experts for graph neural networks

    Sukwon Yun, Kibum Kim, Kanghoon Yoon, and Chanyoung Park. Lte4g: Long-tail experts for graph neural networks. InProceedings of the 31st ACM International Conference on Information & Knowledge Management, pages 2434–2443, 2022

    work page 2022

  5. [5]

    Dgrec: Graph neural network for recommendation with diversified embedding generation

    Liangwei Yang, Shengjie Wang, Yunzhe Tao, Jiankai Sun, Xiaolong Liu, Philip S Yu, and Taiqing Wang. Dgrec: Graph neural network for recommendation with diversified embedding generation. InProceedings of the sixteenth ACM international conference on web search and data mining, pages 661–669, 2023

    work page 2023

  6. [6]

    Graphmasal: A graph-based multi-agent system for adaptive learning.arXiv preprint arXiv:2511.11035, 2025

    Biqing Zeng, Mengquan Liu, and Zongwei Zhen. Graphmasal: A graph-based multi-agent system for adaptive learning.arXiv preprint arXiv:2511.11035, 2025

    work page arXiv 2025

  7. [7]

    Making your first choice: to address cold start problem in medical active learning

    Liangyu Chen, Yutong Bai, Siyu Huang, Yongyi Lu, Bihan Wen, Alan Yuille, and Zongwei Zhou. Making your first choice: to address cold start problem in medical active learning. InMedical Imaging with Deep Learning, pages 496–525. PMLR, 2024

    work page 2024

  8. [8]

    Deuce: Dual-diversity enhancement and uncertainty- awareness for cold-start active learning.Transactions of the Association for Computational Linguistics, 12: 1736–1754, 2024

    Jiaxin Guo, CL Philip Chen, Shuzhen Li, and Tong Zhang. Deuce: Dual-diversity enhancement and uncertainty- awareness for cold-start active learning.Transactions of the Association for Computational Linguistics, 12: 1736–1754, 2024

    work page 2024

  9. [9]

    Semi-supervised batch active learning based on mutual information

    Xia Ji, LingZhu Wang, and XiaoHao Fang. Semi-supervised batch active learning based on mutual information. Applied Intelligence, 55(2):117, 2025. 13 LiveGraph

    work page 2025

  10. [10]

    Few-shot heterogeneous graph learning via cross-domain knowledge transfer

    Qiannan Zhang, Xiaodong Wu, Qiang Yang, Chuxu Zhang, and Xiangliang Zhang. Few-shot heterogeneous graph learning via cross-domain knowledge transfer. InProceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 2450–2460, 2022

    work page 2022

  11. [11]

    Knowledge- enhanced causal reinforcement learning model for interactive recommendation.IEEE Transactions on Multimedia, 26:1129–1142, 2023

    Weizhi Nie, Xin Wen, Jing Liu, Jiawei Chen, Jiancan Wu, Guoqing Jin, Jing Lu, and An-An Liu. Knowledge- enhanced causal reinforcement learning model for interactive recommendation.IEEE Transactions on Multimedia, 26:1129–1142, 2023

    work page 2023

  12. [12]

    Graph meta-reinforcement learning for transferable autonomous mobility-on-demand

    Daniele Gammelli, Kaidi Yang, James Harrison, Filipe Rodrigues, Francisco Pereira, and Marco Pavone. Graph meta-reinforcement learning for transferable autonomous mobility-on-demand. InProceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 2913–2923, 2022

    work page 2022

  13. [13]

    Graph- based re-ranking: Emerging techniques, limitations, and opportunities.arXiv preprint arXiv:2503.14802, 2025

    Md Shahir Zaoad, Niamat Zawad, Priyanka Ranade, Richard Krogman, Latifur Khan, and James Holt. Graph- based re-ranking: Emerging techniques, limitations, and opportunities.arXiv preprint arXiv:2503.14802, 2025

    work page arXiv 2025

  14. [14]

    Personal- ized diversification for neural re-ranking in recommendation

    Weiwen Liu, Yunjia Xi, Jiarui Qin, Xinyi Dai, Ruiming Tang, Shuai Li, Weinan Zhang, and Rui Zhang. Personal- ized diversification for neural re-ranking in recommendation. In2023 IEEE 39th International Conference on Data Engineering (ICDE), pages 802–815. IEEE, 2023

    work page 2023

  15. [15]

    Enhancing recommendation diversity by re-ranking with large language models

    Diego Carraro and Derek Bridge. Enhancing recommendation diversity by re-ranking with large language models. ACM Transactions on Recommender Systems, 4(2):1–40, 2025

    work page 2025

  16. [16]

    Sgccl: Siamese graph contrastive con- sensus learning for personalized recommendation

    Boyu Li, Ting Guo, Xingquan Zhu, Qian Li, Yang Wang, and Fang Chen. Sgccl: Siamese graph contrastive con- sensus learning for personalized recommendation. InProceedings of the sixteenth ACM international conference on web search and data mining, pages 589–597, 2023

    work page 2023

  17. [17]

    Generative-contrastive graph learning for recommendation

    Yonghui Yang, Zhengwei Wu, Le Wu, Kun Zhang, Richang Hong, Zhiqiang Zhang, Jun Zhou, and Meng Wang. Generative-contrastive graph learning for recommendation. InProceedings of the 46th international ACM SIGIR Conference on Research and Development in Information Retrieval, pages 1117–1126, 2023

    work page 2023

  18. [18]

    A comprehensive survey on automatic knowledge graph construction.ACM Computing Surveys, 56(4):1–62, 2023

    Lingfeng Zhong, Jia Wu, Qian Li, Hao Peng, and Xindong Wu. A comprehensive survey on automatic knowledge graph construction.ACM Computing Surveys, 56(4):1–62, 2023

    work page 2023

  19. [19]

    Intelligent tutoring for large-scale personalized programming learning based on knowledge graph

    Ying Li, Jincheng Qiu, Runze Yang, Tongyu Zhu, Hao Sheng, ShiJie Gui, and Yu Liang. Intelligent tutoring for large-scale personalized programming learning based on knowledge graph. In2023 IEEE Frontiers in Education Conference (FIE), pages 1–5. IEEE, 2023

    work page 2023

  20. [20]

    Enhancing the recommendation of learning resources for learners via an advanced knowledge graph.Applied Sciences, 15(8): 4204, 2025

    Chao Duan, Jin Yang, Qiaoling Cui, Wenlong Zhang, Xuelian Wan, and Mingyan Zhang. Enhancing the recommendation of learning resources for learners via an advanced knowledge graph.Applied Sciences, 15(8): 4204, 2025

    work page 2025

  21. [21]

    Research on knowledge graph recommendation method for online education

    Jin Yang, Chao Duan, Zhaozhuan Zeng, Yumeng Liu, Jingjing Bai, and Mingyan Zhang. Research on knowledge graph recommendation method for online education. In2024 4th International Conference on Educational Technology (ICET), pages 59–64. IEEE, 2024

    work page 2024

  22. [22]

    A personalized educational resource recommendation method based on graph neural network

    Xiangting Gao, Xingying Han, and Mario Di Nardo. A personalized educational resource recommendation method based on graph neural network. InProceedings of the 2025 International Conference on Artificial Intelligence and Educational Systems, pages 291–296, 2025

    work page 2025

  23. [23]

    Dynamic knowledge selector and evaluator for recommendation with knowledge graph

    Feng Xia and Zhifei Hu. Dynamic knowledge selector and evaluator for recommendation with knowledge graph. arXiv preprint arXiv:2502.15623, 2025

    work page arXiv 2025

  24. [24]

    Kg4ex: An explainable knowledge graph-based approach for exercise recommendation

    Quanlong Guan, Fang Xiao, Xinghe Cheng, Liangda Fang, Ziliang Chen, Guanliang Chen, and Weiqi Luo. Kg4ex: An explainable knowledge graph-based approach for exercise recommendation. InProceedings of the 32nd ACM international conference on information and knowledge management, pages 597–607, 2023

    work page 2023

  25. [25]

    R-ckgat: a recommendation algorithm based on scientific fitness knowledge graph.Scientific Reports, 15(1):18910, 2025

    Zhitong Liu, Shutong Du, Shouxin Zong, and Bingyu Pan. R-ckgat: a recommendation algorithm based on scientific fitness knowledge graph.Scientific Reports, 15(1):18910, 2025

    work page 2025

  26. [26]

    Knowledge-adaptive contrastive learning for recommendation

    Hao Wang, Yao Xu, Cheng Yang, Chuan Shi, Xin Li, Ning Guo, and Zhiyuan Liu. Knowledge-adaptive contrastive learning for recommendation. InProceedings of the sixteenth ACM international conference on web search and data mining, pages 535–543, 2023

    work page 2023

  27. [27]

    Design knowledge graph-based recommender system for engineering pedagogy

    Yan Li and Zuyang Ma. Design knowledge graph-based recommender system for engineering pedagogy. In2024 12th International Conference on Information and Education Technology (ICIET), pages 58–61. IEEE, 2024. 14 LiveGraph

    work page 2024

  28. [28]

    Graph-based cognitive diagnosis for intelligent tutoring systems.Knowledge-Based Systems, 253:109547, 2022

    Yu Su, Zeyu Cheng, Jinze Wu, Yanmin Dong, Zhenya Huang, Le Wu, Enhong Chen, Shijin Wang, and Fei Xie. Graph-based cognitive diagnosis for intelligent tutoring systems.Knowledge-Based Systems, 253:109547, 2022

    work page 2022

  29. [29]

    Zhen-Yu Cao, Feng Lin, and Chun Feng. Interpretation of course conceptual structure and student self-efficacy: an integrated strategy of knowledge graphs with item response modeling.BMC Medical Education, 24(1):563, 2024

    work page 2024

  30. [30]

    Multivariate knowledge tracking based on graph neural network in assistments.IEEE Transactions on Learning Technologies, 17:32–43, 2023

    Zhenchang Xia, Nan Dong, Jia Wu, and Chuanguo Ma. Multivariate knowledge tracking based on graph neural network in assistments.IEEE Transactions on Learning Technologies, 17:32–43, 2023

    work page 2023

  31. [31]

    Sgkt: Session graph-based knowledge tracing for student performance prediction.Expert Systems with Applications, 206:117681, 2022

    Zhengyang Wu, Li Huang, Qionghao Huang, Changqin Huang, and Yong Tang. Sgkt: Session graph-based knowledge tracing for student performance prediction.Expert Systems with Applications, 206:117681, 2022

    work page 2022

  32. [32]

    Sthkt: Spatiotemporal knowledge tracing with topological hawkes process.Expert Systems with Applications, 259:125248, 2025

    Shuting Li, Shuanghong Shen, Yu Su, Xinjie Sun, Junyu Lu, Qi Mo, Zhenyi Wu, and Qi Liu. Sthkt: Spatiotemporal knowledge tracing with topological hawkes process.Expert Systems with Applications, 259:125248, 2025

    work page 2025

  33. [33]

    Hierarchical uncertainty-aware graph neural network.arXiv preprint arXiv:2504.19820, 2025

    Yoonhyuk Choi, Jiho Choi, Taewook Ko, and Chong-Kwon Kim. Hierarchical uncertainty-aware graph neural network.arXiv preprint arXiv:2504.19820, 2025

    work page arXiv 2025

  34. [34]

    Clst: Cold-start mitigation in knowledge tracing by aligning a generative language model as a students’ knowledge tracer.arXiv preprint arXiv:2406.10296, 2024

    Heeseok Jung, Jaesang Yoo, Yohaan Yoon, and Yeonju Jang. Clst: Cold-start mitigation in knowledge tracing by aligning a generative language model as a students’ knowledge tracer.arXiv preprint arXiv:2406.10296, 2024

    work page arXiv 2024

  35. [35]

    Mitigating cold- start problems in knowledge tracing with large language models: An attribute-aware approach

    Yuxiang Guo, Shuanghong Shen, Qi Liu, Zhenya Huang, Linbo Zhu, Yu Su, and Enhong Chen. Mitigating cold- start problems in knowledge tracing with large language models: An attribute-aware approach. InProceedings of the 33rd ACM International Conference on Information and Knowledge Management, pages 727–736, 2024

    work page 2024

  36. [36]

    Ken utilization layer: Hebbian replay within a student’s ken for adaptive knowledge tracing.arXiv preprint arXiv:2507.00032, 2025

    Grey Kuling and Marinka Zitnik. Ken utilization layer: Hebbian replay within a student’s ken for adaptive knowledge tracing.arXiv preprint arXiv:2507.00032, 2025

    work page arXiv 2025

  37. [37]

    How to measure uncertainty in uncertainty sampling for active learning.Machine Learning, 111(1):89–122, 2022

    Vu-Linh Nguyen, Mohammad Hossein Shaker, and Eyke Hüllermeier. How to measure uncertainty in uncertainty sampling for active learning.Machine Learning, 111(1):89–122, 2022

    work page 2022

  38. [38]

    Active learning with uncertainty-guided sample selection for pet/ct tumor segmentation

    Mahsa Vali, Hossein Arabi, and Katarzyna Bozek. Active learning with uncertainty-guided sample selection for pet/ct tumor segmentation. In2025 IEEE International Conference on Imaging Systems and Techniques (IST), pages 1–6. IEEE, 2025

    work page 2025

  39. [39]

    Uncertainty-guided active learning with theoretical bounds on label complexity.ACM Transactions on Probabilistic Machine Learning, 2025

    Deepesh Agarwal and Balasubramaniam Natarajan. Uncertainty-guided active learning with theoretical bounds on label complexity.ACM Transactions on Probabilistic Machine Learning, 2025

    work page 2025

  40. [40]

    Graph active learning at subgraph granularity

    Yunqi Cao, Ziming Wang, and Haopeng Chen. Graph active learning at subgraph granularity. In2023 IEEE 35th International Conference on Tools with Artificial Intelligence (ICTAI), pages 578–585. IEEE, 2023

    work page 2023

  41. [41]

    Delta: Dual consistency delving with topological uncertainty for active graph domain adaptation.arXiv preprint arXiv:2409.08946, 2024

    Pengyun Wang, Yadi Cao, Chris Russell, Yanxin Shen, Junyu Luo, Ming Zhang, Siyu Heng, and Xiao Luo. Delta: Dual consistency delving with topological uncertainty for active graph domain adaptation.arXiv preprint arXiv:2409.08946, 2024

    work page arXiv 2024

  42. [42]

    Probing classifiers are unreliable for concept removal and detection.Advances in Neural Information Processing Systems, 35:17994–18008, 2022

    Abhinav Kumar, Chenhao Tan, and Amit Sharma. Probing classifiers are unreliable for concept removal and detection.Advances in Neural Information Processing Systems, 35:17994–18008, 2022

    work page 2022

  43. [43]

    Fairdiverse: A comprehensive toolkit for fairness-and diversity-aware information retrieval

    Chen Xu, Zhirui Deng, Clara Rus, Xiaopeng Ye, Yuanna Liu, Jun Xu, Zhicheng Dou, Ji-Rong Wen, and Maarten de Rijke. Fairdiverse: A comprehensive toolkit for fairness-and diversity-aware information retrieval. InProceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 3540–3550, 2025

    work page 2025

  44. [44]

    Submodular optimization: Variants, theory and applications

    Yanhui Zhu. Submodular optimization: Variants, theory and applications. InProceedings of the 33rd ACM International Conference on Information and Knowledge Management, pages 5503–5506, 2024

    work page 2024

  45. [45]

    Diversity-aware deep ranking network for recommendation

    Zihong Wang, Yingxia Shao, Jiyuan He, Jinbao Liu, Shitao Xiao, Tao Feng, and Ming Liu. Diversity-aware deep ranking network for recommendation. InProceedings of the 32nd ACM International Conference on Information and Knowledge Management, pages 2564–2573, 2023

    work page 2023

  46. [46]

    Nr4der: Neural re-ranking for diversified exercise recommendation

    Xinghe Cheng, Xufang Zhou, Liangda Fang, Chaobo He, Yuyu Zhou, Weiqi Luo, Zhiguo Gong, and Quanlong Guan. Nr4der: Neural re-ranking for diversified exercise recommendation. InProceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 1738–1747, 2025. 15 LiveGraph

    work page 2025

  47. [47]

    Lion: A gpu-accelerated online serving system for web-scale recommendation at baidu

    Hao Liu, Qian Gao, Xiaochao Liao, Guangxing Chen, Hao Xiong, Silin Ren, Guobao Yang, and Zhiwei Zha. Lion: A gpu-accelerated online serving system for web-scale recommendation at baidu. InProceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 3388–3397, 2022

    work page 2022

  48. [48]

    Personalized learning path recommendation using graph reinforcement learning.Procedia Computer Science, 258:3480–3489, 2025

    Subhajit Haldar, Souvik Sengupta, and Asit Kumar Das. Personalized learning path recommendation using graph reinforcement learning.Procedia Computer Science, 258:3480–3489, 2025

    work page 2025

  49. [49]

    Reinforcement learning in education: A literature review

    Bisni Fahad Mon, Asma Wasfi, Mohammad Hayajneh, Ahmad Slim, and Najah Abu Ali. Reinforcement learning in education: A literature review. InInformatics, volume 10, page 74. MDPI, 2023

    work page 2023

  50. [50]

    When maml can adapt fast and how to assist when it cannot

    Sébastien Arnold, Shariq Iqbal, and Fei Sha. When maml can adapt fast and how to assist when it cannot. In International conference on artificial intelligence and statistics, pages 244–252. PMLR, 2021

    work page 2021

  51. [51]

    Model-agnostic meta-learning for fast text-dependent speaker embedding adaptation.IEEE/ACM Transactions on Audio, Speech, and Language Processing, 31:1866–1876, 2023

    Weiwei Lin and Man-Wai Mak. Model-agnostic meta-learning for fast text-dependent speaker embedding adaptation.IEEE/ACM Transactions on Audio, Speech, and Language Processing, 31:1866–1876, 2023

    work page 2023

  52. [52]

    On the convergence theory of meta reinforcement learning with personalized policies.arXiv preprint arXiv:2209.10072, 2022

    Haozhi Wang, Qing Wang, Yunfeng Shao, Dong Li, Jianye Hao, and Yinchuan Li. On the convergence theory of meta reinforcement learning with personalized policies.arXiv preprint arXiv:2209.10072, 2022

    work page arXiv 2022

  53. [53]

    A theoretical understanding of gradient bias in meta-reinforcement learning.Advances in Neural Information Processing Systems, 35:31059–31072, 2022

    Bo Liu, Xidong Feng, Jie Ren, Luo Mai, Rui Zhu, Haifeng Zhang, Jun Wang, and Yaodong Yang. A theoretical understanding of gradient bias in meta-reinforcement learning.Advances in Neural Information Processing Systems, 35:31059–31072, 2022

    work page 2022

  54. [54]

    Curricular subgoals for inverse reinforcement learning.IEEE Transactions on Intelligent Transportation Systems, 2025

    Shunyu Liu, Yunpeng Qing, Shuqi Xu, Hongyan Wu, Jiangtao Zhang, Jingyuan Cong, Tianhao Chen, Yun-Fu Liu, and Mingli Song. Curricular subgoals for inverse reinforcement learning.IEEE Transactions on Intelligent Transportation Systems, 2025

    work page 2025

  55. [55]

    Exact: A meta-learning framework for precise exercise segmentation in physical therapy

    Hanchen David Wang, Siwoo Bae, Xutong Sun, Yashvitha Thatigotla, and Meiyi Ma. Exact: A meta-learning framework for precise exercise segmentation in physical therapy. InProceedings of the ACM/IEEE 16th Interna- tional Conference on Cyber-Physical Systems (with CPS-IoT Week 2025), pages 1–11, 2025

    work page 2025

  56. [56]

    semg motion classification via few-shot learning with applications to sports science.Authorea Preprints, 2023

    Mert Ergeneci, Erkan Bayram, Ayberk yarkin Yildiz, Daryl Carter, and Panagiotis Kosmas. semg motion classification via few-shot learning with applications to sports science.Authorea Preprints, 2023

    work page 2023

  57. [57]

    A survey of knowledge graph reasoning on graph types: Static, dynamic, and multi-modal

    Ke Liang, Lingyuan Meng, Meng Liu, Yue Liu, Wenxuan Tu, Siwei Wang, Sihang Zhou, Xinwang Liu, Fuchun Sun, and Kunlun He. A survey of knowledge graph reasoning on graph types: Static, dynamic, and multi-modal. IEEE Transactions on Pattern Analysis and Machine Intelligence, 46(12):9456–9478, 2024

    work page 2024

  58. [58]

    Ivgae: Handling incomplete heterogeneous data with a variational graph autoencoder.arXiv preprint arXiv:2511.22116, 2025

    Youran Zhou, Mohamed Reda Bouadjenek, et al. Ivgae: Handling incomplete heterogeneous data with a variational graph autoencoder.arXiv preprint arXiv:2511.22116, 2025

    work page arXiv 2025

  59. [59]

    Variational graph autoencoder with adversarial mutual information learning for network representation learning.ACM Transactions on Knowledge Discovery from Data, 17(3):1–18, 2023

    Dongjie Li, Dong Li, and Guang Lian. Variational graph autoencoder with adversarial mutual information learning for network representation learning.ACM Transactions on Knowledge Discovery from Data, 17(3):1–18, 2023

    work page 2023

  60. [60]

    Denoising variational graph of graphs auto-encoder for predicting structured entity interactions.IEEE Transactions on Knowledge and Data Engineering, 36(3):1016–1029, 2023

    Han Chen, Hanchen Wang, Hongmei Chen, Ying Zhang, Wenjie Zhang, and Xuemin Lin. Denoising variational graph of graphs auto-encoder for predicting structured entity interactions.IEEE Transactions on Knowledge and Data Engineering, 36(3):1016–1029, 2023

    work page 2023

  61. [61]

    Introducing variational autoencoders to high school students

    Zhuoyue Lyu, Safinah Ali, and Cynthia Breazeal. Introducing variational autoencoders to high school students. In Proceedings of the AAAI conference on artificial intelligence, volume 36, pages 12801–12809, 2022

    work page 2022

  62. [62]

    Vqgraph: Rethinking graph representation space for bridging gnns and mlps.arXiv preprint arXiv:2308.02117, 2023

    Ling Yang, Ye Tian, Minkai Xu, Zhongyi Liu, Shenda Hong, Wei Qu, Wentao Zhang, Bin Cui, Muhan Zhang, and Jure Leskovec. Vqgraph: Rethinking graph representation space for bridging gnns and mlps.arXiv preprint arXiv:2308.02117, 2023

    work page arXiv 2023

  63. [63]

    Graph-based unsupervised disentangled representation learning via multimodal large language models.Advances in Neural Information Processing Systems, 37:103101–103130, 2024

    Baao Xie, Qiuyu Chen, Yunnan Wang, Zequn Zhang, Xin Jin, and Wenjun Zeng. Graph-based unsupervised disentangled representation learning via multimodal large language models.Advances in Neural Information Processing Systems, 37:103101–103130, 2024

    work page 2024

  64. [64]

    Sebot: Structural entropy guided multi-view contrastive learning for social bot detection

    Yingguang Yang, Qi Wu, Buyun He, Hao Peng, Renyu Yang, Zhifeng Hao, and Yong Liao. Sebot: Structural entropy guided multi-view contrastive learning for social bot detection. InProceedings of the 30th ACM SIGKDD conference on knowledge discovery and data mining, pages 3841–3852, 2024. 16 LiveGraph

    work page 2024

  65. [65]

    Boosting bot detection via heterophily-aware representation learning and prototype-guided cluster discovery

    Buyun He, Xiaorui Jiang, Qi Wu, Hao Liu, Yingguang Yang, and Yong Liao. Boosting bot detection via heterophily-aware representation learning and prototype-guided cluster discovery. InProceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V . 2, pages 860–871, 2025

    work page 2025

  66. [66]

    Learning knowledge-diverse experts for long-tailed graph classification.ACM Transactions on Knowledge Discovery from Data, 19(2):1–24, 2025

    Zhengyang Mao, Wei Ju, Siyu Yi, Yifan Wang, Zhiping Xiao, Qingqing Long, Nan Yin, Xinwang Liu, and Ming Zhang. Learning knowledge-diverse experts for long-tailed graph classification.ACM Transactions on Knowledge Discovery from Data, 19(2):1–24, 2025

    work page 2025

  67. [67]

    Deep knowledge tracing.Advances in neural information processing systems, 28, 2015

    Chris Piech, Jonathan Bassen, Jonathan Huang, Surya Ganguli, Mehran Sahami, Leonidas J Guibas, and Jascha Sohl-Dickstein. Deep knowledge tracing.Advances in neural information processing systems, 28, 2015

    work page 2015

  68. [68]

    Addressing the assessment challenge with an online system that tutors as it assesses.User modeling and user-adapted interaction, 19(3):243–266, 2009

    Mingyu Feng, Neil Heffernan, and Kenneth 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

    work page 2009

  69. [69]

    Dynamic key-value memory networks for knowledge tracing

    Jiani Zhang, Xingjian Shi, Irwin King, and Dit-Yan Yeung. Dynamic key-value memory networks for knowledge tracing. InProceedings of the 26th international conference on World Wide Web, pages 765–774, 2017

    work page 2017

  70. [70]

    Context-aware attentive knowledge tracing

    Aritra Ghosh, Neil Heffernan, and Andrew S Lan. Context-aware attentive knowledge tracing. InProceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining, pages 2330–2339, 2020

    work page 2020

  71. [71]

    simplekt: a simple but tough-to-beat baseline for knowledge tracing.arXiv preprint arXiv:2302.06881, 2023

    Zitao Liu, Qiongqiong Liu, Jiahao Chen, Shuyan Huang, and Weiqi Luo. simplekt: a simple but tough-to-beat baseline for knowledge tracing.arXiv preprint arXiv:2302.06881, 2023

    work page arXiv 2023

  72. [72]

    Exercise recommendation based on knowledge concept prediction.Knowledge-Based Systems, 210:106481, 2020

    Zhengyang Wu, Ming Li, Yong Tang, and Qingyu Liang. Exercise recommendation based on knowledge concept prediction.Knowledge-Based Systems, 210:106481, 2020

    work page 2020

  73. [73]

    Meta multi-agent exercise recommendation: A game application perspective

    Fei Liu, Xuegang Hu, Shuochen Liu, Chenyang Bu, and Le Wu. Meta multi-agent exercise recommendation: A game application perspective. InProceedings of the 29th ACM SIGKDD conference on knowledge discovery and data mining, pages 1441–1452, 2023. A Theoretical Analysis and Proofs In this section, we provide the formal mathematical foundations for the LiveGra...

    work page 2023