Pith. sign in

REVIEW 5 major objections 4 minor 1 cited by

Leveraging Large Language Models for Effective Label-free Node Classification in Text-Attributed Graphs

T0 review · 5 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a GNN-driven active self-training loop, Locle, extracts more accuracy from a fixed LLM query budget than LLM-GNN by selecting critical nodes and refining noisy LLM labels with graph rewiring.

desk verdict Locle is a credible, well-engineered LLM+GNN pipeline whose main claim is probably right but is under-supported by the missing validation protocol and a couple of fixable presentation errors. read the letter →

arxiv 2412.11983 v3 pith:FUCVWSRX submitted 2024-12-16 cs.LG cs.AI

Taiyan Zhang , Renchi Yang , Yurui Lai , Mingyu Yan , Xiaochun Ye , Dongrui Fan This is my paper · ORCID
classification cs.LGcs.AI
keywords label-freenodeclassificationtext-attributedgraphslargelanguagemodelsgraphneuralnetworksactivelearningself-trainingpseudo-labelingrewiring
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

Locle is a framework for classifying nodes in text-attributed graphs without any ground-truth labels, using only a small budget of queries to a large language model. The paper argues that the standard LLM-GNN pipeline wastes this budget: it annotates once, ignores graph structure during selection, and lets LLM label noise poison the downstream GNN. Locle instead runs an iterative self-training loop in which a GNN picks the most informative nodes, the LLM labels only the uncertain ones, and a rewired graph topology refines those labels. On five benchmarks the authors report consistent accuracy gains over state-of-the-art baselines under identical LLM query budgets, including an 8.08% improvement on DBLP at a cost below one cent. If correct, this would make LLM-assisted, label-free node classification substantially cheaper and more practical.

What carries the argument

The load-bearing mechanism is the iterative interplay between two signal sources. Stage I builds a $T$-truncated graph-smoothed representation $\mathbf{H} = \sum_{t=0}^T (1-\alpha)\alpha^t \tilde{\mathbf{A}}^t \mathbf{X}$ and runs subspace clustering on it; Lemma 4.1 shows that spectral clustering of the self-expressive matrix $\mathbf{S} = \mathbf{U}\mathbf{U}^\top$ equals $K$-means on the left singular vectors $\mathbf{U}$, so the active node set is cheap to extract. In each self-training round, an ensemble of past GNN predictions is scored by label entropy $LE(v_i)$ and label disharmonicity $LH(v_i)$, the $\ell^2$ norm of the Dirichlet-energy gradient at the node; the most confident nodes become pseudo-labels and the least confident are sent to the LLM. For those uncertain nodes, Locle rewires the graph by optimizing Dirichlet energy with respect to the Laplacian, giving a closed-form adjacency from $\mathbf{H}\mathbf{H}^\top$, then removes low-weight edges, adds edges from labeled to unlabeled nodes, and blends the GNN prediction on the rewired graph with the LLM annotation by rank-based confidence. These two lemmas and the two selection metrics carry the whole argument: they convert graph structure into a label-refinement signal that compensates for LLM noise.

What would settle it

Run Locle on the same five datasets with all hyperparameters chosen by a documented validation split and no test-label access at any point; if the reported accuracy margins over LLM-GNN shrink or disappear, the central claim of generalizable cost-effective gain is refuted.

Watch

Extended reading notes

Core claim

The central claim is that label noise from LLMs, rather than query cost, is the main obstacle to label-free node classification, and that a GNN can actively manage that noise. Locle's first stage selects a small initial annotation set by subspace clustering on smoothed GNN representations; the second stage repeatedly uses label entropy and a new label disharmonicity measure to separate high-confidence nodes, whose GNN predictions become pseudo-labels, from low-confidence nodes, which are sent to the LLM, then refines the LLM outputs by comparing them with predictions on a graph rewired to suppress noisy edges. The paper reports that this pipeline outperforms LLM-GNN and its active-selection variants on Cora, Citeseer, Pubmed, WikiCS, and DBLP for GCN, GAT, and GCNII backbones. It also reports that with GPT-3.5-turbo, Locle can match or beat the same pipeline using stronger GPT-4 models, implying that with the right selection and refinement, the LLM stops being the performance bottleneck and the graph itself supplies most of the supervision.

Load-bearing premise

The load-bearing premise is that the per-dataset settings reported for Locle were chosen without looking at test labels; the paper gives no validation procedure, so if those values were picked to maximize test accuracy on the five benchmarks, the reported gains might not generalize.

Editorial extensions

If this is right

  • Under the same LLM query budget, Locle reports higher accuracy than every LLM-GNN variant on all five datasets for all three GNN backbones, with the largest headline gain being 9.28 percentage points on Cora with GCN.
  • Because Locle with GPT-3.5-turbo matches or beats Locle with GPT-4 on Cora and Citeseer, the framework implies that the choice of LLM matters less than the refinement loop, so users can spend less on model quality.
  • On Cora, once the budget reaches 350 queries, Locle's accuracy approaches that of the same pipeline trained on ground-truth labels, suggesting the label-free gap can be nearly closed.
  • Cost measurements in the paper put querying DBLP at under one cent and make Locle at least twice as fast as the best LLM-GNN variant, so the approach is affordable at benchmark scale.

Reading between the lines

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

  • Locle's components are not specific to LLM noise: the same entropy-plus-disharmonicity selection and rewired-graph refinement could be applied to human annotations with known error rates or to semi-supervised learning with a handful of clean labels, though the paper tests neither setting.
  • Because the rewiring step deliberately adds edges from labeled to unlabeled nodes, a natural extension is to measure how Locle's margin changes with graph homophily or edge density; the paper reports no such sensitivity analysis.
  • The per-dataset settings listed in the paper's hyperparameter table suggest that a new graph would need its own tuning, and the paper offers no label-free procedure for choosing them; transfer to a new domain would require some heuristic or validation scheme.
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

5 major / 4 minor

Summary. The paper proposes Locle, a label-free node classification framework for text-attributed graphs that uses a limited budget of LLM queries. Locle has three main components: (i) an initial active node selection stage based on subspace clustering of GNN-derived features, (ii) a multi-round self-training scheme that selects informative nodes via label entropy and a newly proposed label disharmonicity metric, and (iii) a hybrid label refinement module that combines LLM annotations with predictions from a GNN trained on a rewired graph. The manuscript reports experiments on five TAG datasets with three GNN backbones (GCN, GAT, GCNII), comparing against 19 baselines, and claims consistent accuracy improvements over the LLM-GNN baseline under the same LLM query budget, with a notable 8.08% accuracy improvement highlighted in the abstract. The paper also includes theoretical analyses connecting the proposed metrics to Dirichlet energy and spectral clustering.

Significance. If the central empirical claim holds, the paper makes a useful practical contribution: it addresses two well-known limitations of LLM-GNN pipelines—cost of labeling and noise in LLM-generated labels—by integrating GNN-based selection and refinement. The scope of the evaluation is broad (19 baselines, 5 datasets, 3 backbones), and the code is promised to be public. However, the current manuscript has several load-bearing inconsistencies and evaluation-protocol gaps that prevent the claims from being accepted as stated. In particular, the hyperparameter selection procedure is opaque, the reported hyperparameters conflict with the mathematical formulation in the main text, and one of the theoretical claims contains a factor error. These issues are fixable in revision, but they are central enough that the paper should not be accepted in its present form.

major comments (5)
  1. [Table 10 / Eq. (3), Eq. (7)] The hyperparameter values in Table 10 are inconsistent with the formulas that define the method. Eq. (3) uses the Neumann series H = sum_{t=0}^T (1-alpha) alpha^t Atilde^t X, which requires |alpha| < 1 for convergence; the closed form in Eq. (2) is derived from Eq. (1) under that assumption. Yet Table 10 lists alpha = 1 or alpha = 1.2 for nearly every dataset/backbone combination. If alpha = 1, Eq. (3) gives the zero matrix (since (1-alpha) = 0), and the ensemble weighting in Eq. (7) is 0/0. If alpha = 1.2, the series diverges. The authors must either correct the equations, correct the table, or explain how the implementation computes these quantities for the reported hyperparameters.
  2. [Table 10 and Section 5.1] The manuscript does not describe any validation split or hyperparameter selection protocol. Table 10 reports a different configuration (epsilon, tau, lambda, delta(+), delta(-), phi, B, alpha, and backbone-specific settings) for every dataset and backbone, yet the text only states that results are averaged over three trials. If these values were selected by comparing test accuracies across configurations, the label-free evaluation is compromised because test labels indirectly influence the choice of the LLM budget B and the allocation ratio epsilon. The authors should specify the tuning procedure, e.g., a held-out validation set or a fixed protocol, and report the resulting validation-based selections.
  3. [Section 4.6.1 / Eq. (9)] The claimed identity LH(v_i) = (1/sqrt(|N(v_i)|)) * ||(L Y^(r))_i||_2 is incorrect by a factor sqrt(|N(v_i)|). Since (L Y)_i = |N(v_i)| * (Y_i - (1/|N(v_i)|) sum_{l in N(v_i)} Y_l), taking the Euclidean norm over classes yields ||(L Y)_i||_2 = |N(v_i)| * LH(v_i). The correct relation is LH(v_i) = (1/|N(v_i)|) * ||(L Y)_i||_2 (or equivalently ||(L Y)_i||_2 = |N(v_i)| * LH(v_i)). The theorem as stated should be corrected; otherwise the theoretical justification for the disharmonicity metric does not hold.
  4. [Abstract and Table 2] The abstract states: "on the DBLP dataset with 14.3k nodes, Locle achieves an 8.08% improvement in accuracy over the state-of-the-art." Table 2 shows that the 8.08% accuracy improvement is on WikiCS with the GCN backbone, while the DBLP improvement under GCN is 2.40%. The abstract should be corrected to attribute the 8.08% figure to the correct dataset.
  5. [Tables 2-4 and Section 5.1] All reported accuracies are averages over three trials, but no standard deviations or statistical significance tests are reported. Given that the central claim is "significantly outperforms state-of-the-art," the absence of variance estimates makes it impossible to assess whether the differences (some as small as 0.01% on Pubmed/GAT in Table 2) are reliable. Please report standard deviations or confidence intervals, or provide a significance test.
minor comments (4)
  1. [Section 4.4.1] The symbol H is used for both the node feature matrix (e.g., in Eq. (3)) and the new graph with adjacency matrix H H^T. This notation conflict makes the description of the rewiring step hard to follow. Please use different symbols, e.g., Z for features and G' for the rewired graph.
  2. [Section 5.2 and Abstract] The paper says "19 baselines" in the abstract and contributions, but Section 5.2 says "seven groups of baseline methods, i.e., 20 competitors." The count should be made consistent; counting the rows in Table 2 gives 19 baseline methods.
  3. [Section 1 / Contribution bullet] The fourth contribution bullet claims "a consistent and remarkable improvement of at least 5% in zero-shot classification accuracy compared to the state of the art in most cases." Table 2 contains many improvements well below 5% (e.g., Pubmed GCN: 1.29%; DBLP GCN: 2.40%; Pubmed GAT: 0.01%). Please adjust the wording to match the actual magnitudes.
  4. [Table 2 / LLM as Predictor] The table omits F1 scores for the LLM-as-Predictor baselines; the text explains this is because LLM outputs are invalid for F1 calculation, but it would be helpful to add a footnote in the table itself.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Locle's reported gains rest on an empirical pipeline, not on a derivation that reduces to its own inputs.

full rationale

I walked the claimed derivation chain (Section 4 methodology, Lemmas 4.1-4.2, and the experimental comparisons in Section 5) and found no step in which a predicted quantity is defined in terms of the target result, or in which a fitted parameter is renamed as a prediction. The active node selection, informative sample selection, and hybrid label refinement modules are all computed from GNN probabilities, LLM annotations, and graph structure; the final accuracy is measured against ground-truth labels that are never used to construct the pseudo-labels or to fit the model. The subspace clustering equivalence in Lemma 4.1 is proved directly from S = U U^T, and Lemma 4.2 is established by a self-contained Cauchy-Schwarz argument, so neither theorem smuggles the conclusion in via citation. The citations to the authors' prior graph-clustering work (e.g., [36]) are not load-bearing: the specific subspace-clustering machinery is also attributed to the external references [37,63], and the relevant mathematical claims are proved in the appendix. The one legitimate concern is that Table 10 reports per-dataset hyperparameter values without describing a validation split or model-selection protocol, which could allow test-set-based tuning to inflate the reported margins. That is a benchmarking-transparency risk, not a circularity of the kind defined here: no equation or fitted value is exhibited as reducing to the target accuracy by construction. The absence of an explicit validation procedure is a reproducibility limitation to be weighed under correctness risk, but it does not make the derivation circular. Accordingly, the appropriate circularity score is 0.

Assumptions & free parameters 8 free parameters · 5 assumptions · 0 invented entities

The framework is empirical; its performance depends on the specific datasets, the Sentence-BERT embeddings, the choice of LLM (GPT-3.5-turbo), and many per-dataset hyperparameters. No new physical or conceptual entities are introduced.

free parameters (8)
  • epsilon (budget allocation ratio) = 0.25-0.5 per dataset
    Determines how many of the B LLM queries are spent in the initial annotation stage vs. self-training rounds; tuned per dataset (Table 10).
  • tau (number of left singular vectors) = 64-256 per dataset
    Truncation rank in subspace clustering for active node selection; tuned per dataset (Table 10).
  • lambda (Tikhonov regularizer) = 5e-5 to 5e-3 per dataset
    Weights the ||L_Ghat||_F^2 term in the Dirichlet energy loss (Eq 14); tuned per dataset (Table 10).
  • delta(-) and delta(+) (rewiring ratios) = 0 to 0.3 per dataset
    Fractions of edges removed and added during graph rewiring (Eq 12); tuned per dataset (Table 10).
  • phi (LLM confidence threshold) = 3 or 5 per dataset
    Minimum confidence score to trust LLM annotation in label refinement (Section 4.4.2); tuned per dataset (Table 10).
  • B (total LLM query budget) = 150-400 per dataset
    Total number of LLM queries allowed; set per dataset and kept equal across all compared methods (Table 10).
  • alpha (feature smoothing coefficient) = 1 or 1.2 per dataset
    Coefficient in the GNN representation approximation (Eq 3); tuned per dataset (Table 10).
  • R (number of self-training rounds) = 5 or 6 per dataset
    Chosen from Figure 4(b) as the value giving best accuracy; not listed in Table 10 but determined per dataset.
assumptions (5)
  • standard math The minimizer of Eq (4) with nuclear norm regularization is S = U U^T, where U is the left singular vectors of the data matrix.
    Invoked in Section 4.2.1 to justify subspace clustering; taken from [37] and proved in prior work.
  • domain assumption The GNN representation in Eq (3) approximates the closed-form solution of the graph Laplacian smoothing objective (Eq 1).
    Used in Section 4.2.1 to compute feature vectors for active node selection; relies on the established interpretation of GCN/APPNP as Laplacian smoothing [43,81].
  • domain assumption LLM annotations with confidence scores are accurate enough to serve as training labels, and post-filtering removes low-quality ones.
    Core to both Stage I and the label refinement module; the paper measures LLM annotation accuracy at about 75% (Table 13), so this assumption is partially quantified.
  • ad hoc to paper The rewired graph, built from H H^T with positive weights, improves label propagation for uncertain nodes.
    Proposed in Section 4.4.1; there is no guarantee that adding edges to similar labeled nodes helps, and the ablation shows it does on average.
  • domain assumption Label disharmonicity as defined in Eq 9 is a reliable indicator of label uncertainty.
    Used for informative sample selection; the paper connects it to Dirichlet energy but with a factor error (Section 4.6.1).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Leveraging Large Language Models for Effective Label-free Node Classification in Text-Attributed Graphs." pith.science (2026). https://pith.science/paper/FUCVWSRX

@misc{pith2026241211983,
  author       = {Pith},
  title        = {Pith review of: Leveraging Large Language Models for Effective Label-free Node Classification in Text-Attributed Graphs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FUCVWSRX}},
  note         = {Machine review of arXiv:2412.11983}
}
read the original abstract

Graph neural networks (GNNs) have become the preferred models for node classification in graph data due to their robust capabilities in integrating graph structures and attributes. However, these models heavily depend on a substantial amount of high-quality labeled data for training, which is often costly to obtain. With the rise of large language models (LLMs), a promising approach is to utilize their exceptional zero-shot capabilities and extensive knowledge for node labeling. Despite encouraging results, this approach either requires numerous queries to LLMs or suffers from reduced performance due to noisy labels generated by LLMs. To address these challenges, we introduce Locle, an active self-training framework that does Label-free node Classification with LLMs cost-Effectively. Locle iteratively identifies small sets of "critical" samples using GNNs and extracts informative pseudo-labels for them with both LLMs and GNNs, serving as additional supervision signals to enhance model training. Specifically, Locle comprises three key components: (i) an effective active node selection strategy for initial annotations; (ii) a careful sample selection scheme to identify "critical" nodes based on label disharmonicity and entropy; and (iii) a label refinement module that combines LLMs and GNNs with a rewired topology. Extensive experiments on five benchmark text-attributed graph datasets demonstrate that Locle significantly outperforms state-of-the-art methods under the same query budget to LLMs in terms of label-free node classification. Notably, on the DBLP dataset with 14.3k nodes, Locle achieves an 8.08% improvement in accuracy over the state-of-the-art at a cost of less than one cent. Our code is available at https://github.com/HKBU-LAGAS/Locle.

Figures

Figures reproduced from arXiv: 2412.11983 by the authors.

Figure 1
Figure 1. Varying #labeled nodes. expert knowledge, significant human efforts, and potential biases in annotation, particularly for large-scale graphs comprising millions of nodes and a sheer volume of textual data [10, 56]. In light of the superb comprehension and reasoning abilities of large language models (LLMs) in dealing with textual data, LLMs have been employed as a powerful tool for analyzing TAGs. As manifested in [… view at source ↗
Figure 2
Figure 2. Pipeline of Our Proposed Locle where Ω(𝑺) signifies a regularization term introduced to impose additional structure constraints on 𝑺. A popular choice for Ω(𝑺) is the nuclear norm ∥𝑺 ∥∗, which is to promote the low-rankness of 𝑺 [37]. As such, if we let 𝑼 be the left singular vectors of 𝑿, the minimizer of Eq. (4) is uniquely given by 𝑺 = 𝑼 𝑼 ⊤ [37]. Lemma 4.1. The spectral clustering of 𝑺 with 𝐾 desired clusters is… view at source ↗
Figure 3
Figure 3. Varying 𝐵 in Locle. three variants and Locle, which exhibit the efficacy of our pro￾posed techniques in Locle. In particular, the variant, Locle w/o initial active node selection, produces the lowest accuracy results, indicating the importance of selecting representative nodes as the initial training samples. Analysis of different Active Node Selection methods. To verify the efficacy of our subspace clustering techn… view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: Varying parameters in Locle. B.2 Datasets and Metrics Cora, Citeseer and Pubmed are three widely used datasets in the GNN community. Each node indicates a paper, and the edges indi￾cate the citation relation. We get the raw text of each paper from [6]. Wiki-CS consists…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. When LLMs meet open-world graph learning: a new perspective for unlabeled data uncertainty

    cs.LG 2025-05 reject novelty 6.0 of 10

    OGA combines prototype-based unknown-class rejection with LLM-generated, structure-guided annotations so text-attributed graphs can be retrained in open-world settings.

Reference graph

Works this paper leans on

84 extracted references · 49 canonical work pages · cited by 1 Pith paper

  1. [1]

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Floren- cia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774 (2023)

    arXiv 2023

  2. [2]

    Haoli Bai, Zhuangbin Chen, Michael R Lyu, Irwin King, and Zenglin Xu. 2018. Neural relational topic models for scientific article analysis. In CIKM. 27–36

    2018

  3. [3]

    Cécile Bothorel, Juan David Cruz, Matteo Magnani, and Barbora Micenkova. 2015. Clustering attributed graphs: models, measures and methods. Network Science 3, 3 (2015), 408–444

    2015

  4. [4]

    Jonathan Chang and David Blei. 2009. Relational topic models for document networks. In Artificial intelligence and statistics. PMLR, 81–88

    2009

  5. [5]

    Ming Chen, Zhewei Wei, Zengfeng Huang, Bolin Ding, and Yaliang Li. 2020. Simple and deep graph convolutional networks. In ICML. PMLR, 1725–1735

    2020

  6. [6]

    Zhikai Chen, Haitao Mao, Hang Li, Wei Jin, Hongzhi Wen, Xiaochi Wei, Shuaiqiang Wang, Dawei Yin, Wenqi Fan, Hui Liu, et al. 2024. Exploring the poten- tial of large language models (llms) in learning on graphs. SIGKDD Explorations Newsletter 25, 2 (2024), 42–61

    work page 2024

  7. [7]

    Zhikai Chen, Haitao Mao, Hongzhi Wen, Haoyu Han, Wei Jin, Haiyang Zhang, Hui Liu, and Jiliang Tang. 2023. Label-free Node Classification on Graphs with Large Language Models (LLMs). In ICLR

    work page 2023

  8. [8]

    Petr Chunaev. 2020. Community detection in node-attributed social networks: a survey. Computer Science Review 37 (2020), 100286

    work page 2020

Show all 84 references
  1. [9]

    Fan RK Chung. 1997. Spectral graph theory. Vol. 92. American Mathematical Soc

    1997

  2. [10]

    Enyan Dai, Charu Aggarwal, and Suhang Wang. 2021. NRGNN: Learning a Label Noise-Resistant Graph Neural Network on Sparsely and Noisily Labeled Graphs. arXiv:2106.04714 [cs.LG] https://arxiv.org/abs/2106.04714

    2021 arXiv

  3. [11]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. arXiv:1810.04805 [cs.CL] https://arxiv.org/abs/1810.04805

    2019 arXiv

  4. [12]

    Xiaowen Dong, Dorina Thanou, Pascal Frossard, and Pierre Vandergheynst

  5. [13]

    Yuan Fang, Bo-June Hsu, and Kevin Chen-Chuan Chang. 2012. Confidence- aware graph regularization with heterogeneous pairwise features. In Proceedings of the 35th international ACM SIGIR conference on Research and development in information retrieval. 951–960

    2012

  6. [14]

    Chakib Fettal, Lazhar Labiod, and Mohamed Nadif. 2023. Scalable attributed- graph subspace clustering. In AAAI, Vol. 37. 7559–7567

    2023

  7. [15]

    Dongqi Fu and Jingrui He. 2021. Sdg: A simplified and dynamic graph neural net- work. In Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval . 2273–2277

    2021

  8. [16]

    Johannes Gasteiger, Aleksandar Bojchevski, and Stephan Günnemann. 2018. Predict then Propagate: Graph Neural Networks meet Personalized PageRank. In ICLR

    2018

  9. [17]

    C Lee Giles, Kurt D Bollacker, and Steve Lawrence. 1998. CiteSeer: An automatic citation indexing system. In DL. 89–98

    1998

  10. [18]

    Justin Gilmer, Samuel S Schoenholz, Patrick F Riley, Oriol Vinyals, and George E Dahl. 2017. Neural message passing for quantum chemistry. In ICML. PMLR, 1263–1272

    2017

  11. [19]

    Lei Gong, Sihang Zhou, Wenxuan Tu, and Xinwang Liu. 2022. Attributed Graph Clustering with Dual Redundancy Reduction.. In IJCAI. 3015–3021

    2022

  12. [20]

    Lars Hagen and Andrew B Kahng. 1992. New spectral methods for ratio cut partitioning and clustering. IEEE TCAD 11, 9 (1992), 1074–1085

    1992

  13. [21]

    Will Hamilton, Zhitao Ying, and Jure Leskovec. 2017. Inductive representation learning on large graphs. NeurIPS 30 (2017)

    2017

  14. [22]

    Jie Hao and William Zhu. 2023. Deep graph clustering with enhanced feature representations for community detection. Applied Intelligence 53, 2 (2023), 1336– 1349

    2023

  15. [23]

    Xiaoxin He, Xavier Bresson, Thomas Laurent, Adam Perold, Yann LeCun, and Bryan Hooi. 2024. Harnessing Explanations: LLM-to-LM Interpreter for Enhanced Text-Attributed Graph Representation Learning. In ICLR

    2024

  16. [24]

    2012.Matrix analysis

    Roger A Horn and Charles R Johnson. 2012.Matrix analysis. Cambridge university press

    2012

  17. [25]

    Keke Huang, Jing Tang, Juncheng Liu, Renchi Yang, and Xiaokui Xiao. 2023. Node-wise diffusion for scalable graph learning. In TheWebConf. 1723–1733

    2023

  18. [26]

    Xuanwen Huang, Kaiqiao Han, Yang Yang, Dezheng Bao, Quanjin Tao, Ziwei Chai, and Qi Zhu. 2024. Can GNN be Good Adapter for LLMs? arXiv:2402.12984

    2024 arXiv

  19. [27]

    Ming Ji, Yizhou Sun, Marina Danilevsky, Jiawei Han, and Jing Gao. 2010. Graph regularized transductive classification on heterogeneous information networks. In ECMLKDD. Springer, 570–586

    2010

  20. [28]

    Wei Ju, Yifang Qin, Siyu Yi, Zhengyang Mao, Kangjie Zheng, Luchen Liu, Xiao Luo, and Ming Zhang. 2023. Zero-shot node classification with graph contrastive embedding network. TMLR (2023)

    2023

  21. [29]

    Thomas N Kipf and Max Welling. 2016. Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907 (2016)

    2016 arXiv

  22. [30]

    Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Ves Stoyanov, and Luke Zettlemoyer. 2019. BART: De- noising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension. arXiv:1910.13461

    2019 arXiv

  23. [31]

    Quan Li, Tianxiang Zhao, Lingwei Chen, Junjie Xu, and Suhang Wang. 2024. En- hancing Graph Neural Networks with Limited Labeled Data by Actively Distilling Knowledge from Large Language Models. arXiv:2407.13989

    2024 arXiv

  24. [32]

    Yiran Li, Gongyao Guo, Jieming Shi, Renchi Yang, Shiqi Shen, Qing Li, and Jun Luo. 2024. A versatile framework for attributed network clustering via K-nearest neighbor augmentation. The VLDB Journal 33, 6 (2024), 1913–1943

    2024

  25. [33]

    Yuexin Li and Bryan Hooi. 2023. Prompt-based zero-and few-shot node classifi- cation: A multimodal approach. arXiv preprint arXiv:2307.11572 (2023)

    2023 arXiv

  26. [34]

    Yuhan Li, Peisong Wang, Zhixun Li, Jeffrey Xu Yu, and Jia Li. 2024. ZeroG: Inves- tigating Cross-dataset Zero-shot Transferability in Graphs. arXiv:2402.11235

    2024 arXiv

  27. [35]

    Yiran Li, Renchi Yang, and Jieming Shi. 2023. Efficient and Effective Attributed Hypergraph Clustering via K-Nearest Neighbor Augmentation.Proc. ACM Manag. Data 1, 2 (2023), 116:1–116:23

    2023

  28. [36]

    Xiaoyang Lin, Renchi Yang, Haoran Zheng, and Xiangyu Ke. 2024. Spectral Subspace Clustering for Attributed Graphs. arXiv preprint arXiv:2411.11074 (2024)

    2024 arXiv

  29. [37]

    Guangcan Liu, Zhouchen Lin, and Yong Yu. 2010. Robust subspace segmentation by low-rank representation. In ICML. 663–670

    2010

  30. [38]

    Hao Liu, Jiarui Feng, Lecheng Kong, Ningyue Liang, Dacheng Tao, Yixin Chen, and Muhan Zhang. 2024. One for All: Towards Training One Graph Model for All Classification Tasks. arXiv:2310.00149

    2024 arXiv

  31. [39]

    Pengfei Liu, Weizhe Yuan, Jinlan Fu, Zhengbao Jiang, Hiroaki Hayashi, and Graham Neubig. 2023. Pre-train, prompt, and predict: A systematic survey of prompting methods in natural language processing. Comput. Surveys 55, 9 (2023), 1–35

    2023

  32. [40]

    Yue Liu, Wenxuan Tu, Sihang Zhou, Xinwang Liu, Linxuan Song, Xihong Yang, and En Zhu. 2022. Deep graph clustering via dual correlation reduction. In AAAI, Vol. 36. 7603–7611

    2022

  33. [41]

    Yue Liu, Jun Xia, Sihang Zhou, Xihong Yang, Ke Liang, Chenchen Fan, Yan Zhuang, Stan Z Li, Xinwang Liu, and Kunlun He. 2022. A Survey of Deep Graph Clustering: Taxonomy, Challenge, Application, and Open Resource.arXiv preprint arXiv:2211.12875 (2022)

    2022 arXiv

  34. [42]

    Jiaqi Ma, Ziqiao Ma, Joyce Chai, and Qiaozhu Mei. 2022. Partition-based active learning for graph neural networks. arXiv preprint arXiv:2201.09391 (2022). 10 Leveraging Large Language Models for Effective Label-free Node Classification in Text-Attributed Graphs SIGIR ’25, July...

    2022 arXiv

  35. [43]

    Yao Ma, Xiaorui Liu, Tong Zhao, Yozen Liu, Jiliang Tang, and Neil Shah. 2020. A Unified View on Graph Neural Networks as Graph Signal Denoising. CIKM (2020)

    2020

  36. [44]

    Zhengyi Ma, Zhicheng Dou, Wei Xu, Xinyu Zhang, Hao Jiang, Zhao Cao, and Ji-Rong Wen. 2021. Pre-training for ad-hoc retrieval: hyperlink is also you need. In CIKM. 1212–1221

    2021

  37. [45]

    Kelong Mao, Xi Xiao, Jieming Zhu, Biao Lu, Ruiming Tang, and Xiuqiang He

  38. [46]

    Andrew Kachites McCallum, Kamal Nigam, Jason Rennie, and Kristie Seymore

  39. [48]

    Péter Mernyei and Cătălina Cangea. 2022. Wiki-CS: A Wikipedia-Based Bench- mark for Graph Neural Networks. arXiv:2007.02901

    2022 arXiv

  40. [49]

    Zhihao Peng, Hui Liu, Yuheng Jia, and Junhui Hou. 2021. Attention-driven graph clustering network. In MM. 935–943

    2021

  41. [50]

    Michael D Plummer and László Lovász. 1986. Matching theory. Elsevier

    1986

  42. [51]

    Vipula Rawte, Amit Sheth, and Amitava Das. 2023. A survey of hallucination in large foundation models. arXiv preprint arXiv:2309.05922 (2023)

    2023 arXiv

  43. [52]

    Nils Reimers and Iryna Gurevych. 2019. Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks. arXiv:1908.10084

    2019 arXiv

  44. [53]

    Boon-Siew Seah, Aixin Sun, and Sourav S Bhowmick. 2018. Killing two birds with one stone: Concurrent ranking of tags and comments of social images. InThe 41st International ACM SIGIR Conference on Research & Development in Information Retrieval. 937–940

    2018

  45. [54]

    Prithviraj Sen, Galileo Namata, Mustafa Bilgic, Lise Getoor, Brian Galligher, and Tina Eliassi-Rad. 2008. Collective classification in network data. AI magazine 29, 3 (2008), 93–93

    2008

  46. [55]

    Claude Elwood Shannon. 1948. A mathematical theory of communication. The Bell system technical journal 27, 3 (1948), 379–423

    1948

  47. [56]

    Victor S Sheng, Foster Provost, and Panagiotis G Ipeirotis. 2008. Get another label? improving data quality and data mining using multiple, noisy labelers. In SIGKDD. 614–622

    2008

  48. [57]

    Jiabin Tang, Yuhao Yang, Wei Wei, Lei Shi, Lixin Su, Suqi Cheng, Dawei Yin, and Chao Huang. 2024. GraphGPT: Graph Instruction Tuning for Large Language Models. arXiv:2310.13023

    2024 arXiv

  49. [58]

    Andrey Nikolayevich Tikhonov. 1977. Solutions of Ill-Posed Problems. VH Winston and Sons (1977)

    1977

  50. [59]

    Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, et al. 2023. Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971 (2023)

    2023 arXiv

  51. [60]

    Wenxuan Tu, Sihang Zhou, Xinwang Liu, Xifeng Guo, Zhiping Cai, En Zhu, and Jieren Cheng. 2021. Deep fusion clustering network. InAAAI, Vol. 35. 9978–9987

    2021

  52. [61]

    A Vaswani. 2017. Attention is all you need. NeurIPS (2017)

    2017

  53. [62]

    Petar Veličković, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Lio, and Yoshua Bengio. 2017. Graph attention networks. arXiv preprint arXiv:1710.10903 (2017)

    2017 arXiv

  54. [63]

    René Vidal. 2011. Subspace clustering. IEEE Signal Processing Magazine 28, 2 (2011), 52–68

    2011

  55. [64]

    Ulrike Von Luxburg. 2007. A tutorial on spectral clustering. Statistics and computing 17 (2007), 395–416

    2007

  56. [65]

    Chun Wang, Shirui Pan, Ruiqi Hu, Guodong Long, Jing Jiang, and Chengqi Zhang

  57. [66]

    Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc V Le, Ed H Chi, Sharan Narang, Aakanksha Chowdhery, and Denny Zhou. 2023. Self-Consistency Improves Chain of Thought Reasoning in Language Models. In ICLR

    2023

  58. [67]

    Zheng Wang, Jialong Wang, Yuchen Guo, and Zhiguo Gong. 2021. Zero-shot node classification with decomposed graph prototype network. InSIGKDD. 1769–1779

    2021

  59. [68]

    Zhihao Wen and Yuan Fang. 2023. Augmenting low-resource text classification with graph-grounded pre-training and prompting. In SIGIR. 506–516

    2023

  60. [69]

    Felix Wu, Amauri Souza, Tianyi Zhang, Christopher Fifty, Tao Yu, and Kilian Weinberger. 2019. Simplifying graph convolutional networks. In ICML. 6861– 6871

    2019

  61. [70]

    Yuexin Wu, Yichong Xu, Aarti Singh, Yiming Yang, and Artur Dubrawski. 2019. Active learning for graph neural networks via node feature propagation. arXiv preprint arXiv:1910.07567 (2019)

    2019 arXiv

  62. [71]

    Qianqian Xie, Yutao Zhu, Jimin Huang, Pan Du, and Jian-Yun Nie. 2021. Graph neural collaborative topic model for citation recommendation. TOIS 40, 3 (2021), 1–30

    2021

  63. [72]

    Keyulu Xu, Chengtao Li, Yonglong Tian, Tomohiro Sonobe, Ken-ichi Kawarabayashi, and Stefanie Jegelka. 2018. Representation learning on graphs with jumping knowledge networks. In ICML. 5453–5462

    2018

  64. [73]

    Xiongfeng Yan, Tinghua Ai, Min Yang, and Xiaohua Tong. 2021. Graph convo- lutional autoencoder model for the shape coding and cognition of buildings in maps. IJGIS 35, 3 (2021), 490–512

    2021

  65. [74]

    Renchi Yang, Jieming Shi, Yin Yang, Keke Huang, Shiqi Zhang, and Xiaokui Xiao. 2021. Effective and scalable clustering on massive attributed graphs. In TheWebConf. 3675–3687

    2021

  66. [75]

    Renchi Yang, Yidu Wu, Xiaoyang Lin, Qichen Wang, Tsz Nam Chan, and Jieming Shi. 2024. Effective Clustering on Large Attributed Bipartite Graphs. In Pro- ceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. 3782–3793

    2024

  67. [76]

    Xihong Yang, Yue Liu, Sihang Zhou, Siwei Wang, Wenxuan Tu, Qun Zheng, Xinwang Liu, Liming Fang, and En Zhu. 2023. Cluster-guided contrastive graph clustering network. In AAAI, Vol. 37. 10834–10842

    2023

  68. [77]

    Michihiro Yasunaga, Jure Leskovec, and Percy Liang. 2022. Linkbert: Pretraining language models with document links. arXiv preprint arXiv:2203.15827 (2022)

    2022 arXiv

  69. [78]

    Delvin Ce Zhang and Hady W Lauw. 2021. Semi-supervised semantic visualiza- tion for networked documents. In ECML PKDD. Springer, 762–778

    2021

  70. [79]

    Jiawei Zhang, Haopeng Zhang, Congying Xia, and Li Sun. 2020. Graph-bert: Only attention is needed for learning graph representations. arXiv preprint arXiv:2001.05140 (2020)

    2020 arXiv

  71. [80]

    Wentao Zhang, Yexin Wang, Zhenbang You, Meng Cao, Ping Huang, Jiulong Shan, Zhi Yang, and Bin Cui. 2021. Rim: Reliable influence-based active learning on graphs. NeurIPS 34 (2021), 27978–27990

    2021

  72. [81]

    answer": <answer_here>,

    Meiqi Zhu, Xiao Wang, Chuan Shi, Houye Ji, and Peng Cui. 2021. Interpreting and Unifying Graph Neural Networks with An Optimization Framework.TheWebConf (2021). 11 SIGIR ’25, July 13–18, 2025, Padua, Italy Taiyan Zhang, Renchi Yang, Yurui Lai, Mingyu Yan, Xiaochun Ye, and Dong...

    2021

  73. [2000]

    Information Retrieval 3 (2000), 127–163

    Automating the construction of internet portals with machine learning. Information Retrieval 3 (2000), 127–163

    2000

  74. [2016]

    IEEE Transactions on Signal Processing 64, 23 (2016), 6160–6173

    Learning Laplacian matrix in smooth graph signal representations. IEEE Transactions on Signal Processing 64, 23 (2016), 6160–6173

    2016

  75. [2019]

    Attributed graph clustering: A deep attentional embedding approach.arXiv preprint arXiv:1906.06532 (2019)

    2019 arXiv

  76. [2020]

    In SIGIR

    Item tagging for information retrieval: A tripartite graph neural network based approach. In SIGIR. 2327–2336

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.