Pith. sign in

REVIEW 2 major objections 6 minor 160 references

Beyond Looking Up, Try Looking Around: Harmonizing Global Structure and Local Consistency in Optimal Transport for Short Text Clustering

T0 review · 2 major / 6 minor · reviewed 2026-07-14 · grok-4.5

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

Pith's one-line read Optimal transport that also looks at neighboring samples yields more reliable pseudo-labels for short-text clustering.

desk verdict Solid engineering fix for OT pseudo-labeling: the bilinear consistency term is new and the gains are real, but the closed-loop attention training is the soft spot the paper never fully diagnoses. read the letter →

arxiv 2607.10548 v1 pith:6BVH7Z75 submitted 2026-07-12 stat.ML cs.LG

Zhihao Yao , Yuxuan Gu , Jixuan Yin , Bo Li This is my paper
classification stat.MLcs.LG
keywords shorttextclusteringoptimaltransportpseudo-labelinginstance-levelattentionsemanticconsistencyEMimbalanced
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

Short-text clustering often relies on optimal transport to assign pseudo-labels that respect the global layout of samples versus clusters. Prior transport methods ignore whether nearby, semantically similar texts receive the same label, so ambiguous samples near decision boundaries can be mislabeled and then poison later training. This paper introduces consistency-aware adaptive optimal transport (CAOT): an instance-level attention network first estimates pairwise semantic similarity, and that similarity matrix is inserted as an extra term into the transport objective. Solving the resulting problem produces soft assignments that simultaneously minimize sample-to-cluster cost and encourage similar samples to share labels. Those labels then supervise both the clustering head and the attention network itself inside an EM loop. On eight standard short-text benchmarks the resulting clusters beat recent OT baselines on most accuracy and NMI scores, and the same pipeline also works for long documents and images.

What carries the argument

Consistency-aware adaptive optimal transport (CAOT): the minimization of sample-to-cluster cost plus entropy and imbalance penalties minus an inner-product term that aligns the transport matrix Q with a learned similarity matrix S (Eq. 3).

What would settle it

Replace the learned attention matrix S_att with a pure cosine matrix (or with random similarities) while keeping every other component fixed; if clustering accuracy on the eight benchmarks no longer improves over plain OT, the claimed benefit of the consistency term disappears.

Watch

Extended reading notes

Core claim

When the classic OT objective is augmented with a neighborhood-consistency term that rewards transport vectors of attention-similar samples to be alike, the recovered pseudo-labels become more accurate and confident than pure sample-to-cluster transport, and the labels can be used to train a joint clustering model that outperforms prior short-text methods.

Load-bearing premise

That the attention similarity matrix, which is itself trained with the same pseudo-labels it helps produce, is a faithful enough proxy for true semantic neighborhoods that the extra consistency term corrects early mistakes rather than locking them in.

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

2 major / 6 minor

Summary. The paper proposes CAOT-STC, an EM-style short-text clustering framework whose E-step solves a consistency-aware adaptive optimal transport problem (Eq. 3). The objective augments the usual sample-to-cluster cost with an entropy term, a soft uniformity penalty Ψ(b) that accommodates class imbalance, and a quadratic semantic-consistency term −ε3⟨S,QQᵀ⟩ where S = S_cos + S_att. S_att is produced by an instance-level attention network G_h that is itself trained with the same CAOT pseudo-labels (L_A, Eq. 10). After a K-means warm-up the three modules (PGM, SSCM, SGM) are jointly optimized. Experiments on eight short-text benchmarks, plus long-text and image transfers, report gains over RSTC, SCPCL and FNSCC; ablations, representation-quality plots, coupling-matrix visualizations and a CV-based rule for ε2 are supplied. Code is released.

Significance. If the claimed local-consistency correction is genuine, the work supplies a practical and scalable improvement to OT-based pseudo-labeling that is especially useful for short, sparse text and for imbalanced label distributions. Strengths that raise the paper above a pure empirical tweak include: (i) an explicit, solvable OT formulation that jointly encodes global structure and neighborhood consistency; (ii) a mini-batch solver whose runtime scales better than full-dataset OT (Tables 9–10); (iii) a simple CV heuristic that removes manual tuning of the imbalance weight ε2; (iv) public code and cross-domain checks on long text and images. These elements make the contribution reusable beyond the eight short-text tables.

major comments (2)
  1. The central claim that −ε3⟨S,QQᵀ⟩ injects genuine sample-to-sample semantic consistency rests on S_att remaining a faithful neighborhood proxy after the K-means warm-up. After Algorithm 1 leaves the warm-up stage, S_att is trained exclusively with CAOT-derived same-cluster sets R_i (Eq. 9) via L_A (Eq. 10). The manuscript never reports an independent diagnostic—e.g., precision/recall of S_att against ground-truth same-class pairs, or correlation of S_att with frozen SBERT cosine neighborhoods—once EM begins. Without such a check it is possible that S_att merely amplifies the warm-up partition, rendering the “local-consistency correction” illusory and leaving open the possibility that the observed gains over RSTC/SCPCL arise from contrastive fine-tuning or the soft uniformity term alone. A short diagnostic table or plot would make the load-bearing assumption falsifiable.
  2. Table 2 shows that OURS is not uniformly superior: on SearchSnippets ACC drops 5.04 points relative to FNSCC, and on GoogleNews-TS ACC drops 3.44 points relative to SCPCL. The abstract and §4.2 claim that the method “outperforms state-of-the-art methods” and ranks first on eleven of sixteen indicators; the mixed results should be acknowledged more carefully and the conditions under which CAOT helps versus hurts should be discussed, otherwise the central empirical claim is overstated.
minor comments (6)
  1. Eq. (3) writes a = 1/n 1_n×1 while Appendix A.1 writes a = 1/N 1_n×1; the symbol N is never defined. Unify notation.
  2. Figure 1 caption and the surrounding text refer to “hexagons” and “blue triangles”; the figure itself is schematic and hard to parse without color. A clearer legend or grayscale-safe markers would help.
  3. In Algorithm 1 the interactive-stage loss is written as Eq. (12) but the text defines it as Eq. (13); fix the cross-reference.
  4. Table 3 “w/o All” still contains S_cos in some rows according to the prose; clarify whether the ablation removes both similarities or only the attention term.
  5. The CV-based mapping for ε2 (Table 4 and §4.5) is useful but the thresholds [0,0.2), [0.2,0.4), … appear chosen post-hoc; a short sensitivity check or leave-one-dataset-out validation would strengthen the claim of automatic configuration.
  6. Typos: “short in modeling” (abstract), “Pseduo-labeling” (Appendix B.1 heading), and inconsistent capitalization of dataset names.

Circularity Check

1 steps flagged · score 2.0 of 10

Ordinary EM self-training loop between CAOT pseudo-labels and S_att; no derivation-by-construction or load-bearing self-citation circularity.

  1. self definitional [§3.2–3.3, Eq. (3), Eq. (9)–(10), Algorithm 1]
    "By minimizing L_A, samples within the same cluster are assigned high weights in S_att, which in turn facilitates CAOT in the subsequent E-step. In the training process, accurate pseudo-labels help establish an accurate S_att (M-step), which in turn facilitates the generation of more reliable pseudo-labels than those in the previous iteration (E-step)."

    After warm-up, the similarity matrix S = S_cos + S_att that appears in the CAOT objective is itself produced by an attention network supervised by the CAOT-derived pseudo-labels ŷ (via R_i and L_A). Thus the “semantic consistency” term −ε3⟨S,QQᵀ⟩ is shaped by the very labels it is supposed to improve; the loop is self-referential by construction once EM begins. The paper mitigates but does not eliminate the dependence (K-means warm-up + independent L_I).

full rationale

This is an empirical method paper proposing an OT objective (Eq. 3) plus an attention network, solved by a standard Taylor+Lagrange iteration (Appendix A) and trained in an EM fashion with a K-means warm-up (Algorithm 1). The only closed loop is the usual self-training pattern: CAOT ŷ define same-cluster sets R_i that supervise L_A, which shapes S_att that re-enters the next CAOT solve. That loop is partially broken by the independent contrastive loss L_I, the warm-up stage that seeds ŷ without CAOT, and external ACC/NMI evaluation against ground-truth labels on eight public benchmarks. No parameter is fitted to a target quantity and then re-presented as a prediction; no uniqueness theorem or ansatz is imported from overlapping authors as an external fact; the solution procedure does not reduce to its inputs by algebraic identity. The mild self-training dependence is therefore ordinary engineering practice rather than circular derivation, warranting only a low score.

Assumptions & free parameters 5 free parameters · 3 assumptions · 2 invented entities

The central empirical claim rests on a small set of free coefficients that control the relative strength of entropy, imbalance and consistency terms, on the modeling assumption that cosine-plus-attention similarity is a valid semantic neighborhood, and on the invented CAOT objective itself. No new physical entities are postulated; the free parameters are ordinary ML hyper-parameters whose values are either fixed or set by a simple CV heuristic.

free parameters (5)
  • ε1 (entropy weight) = 1
    Controls softness of the transport plan; fixed to 1 across all datasets after a small grid search.
  • ε2 (imbalance penalty weight) = 100 / 3.5 / 0.06 / 0.03 by imbalance bin
    Strength of the soft uniformity regularizer Ψ(b); set according to a four-bin CV heuristic that itself depends on an external embedding model (bge-large).
  • ε3 (consistency weight) = 25
    Weight of the novel −⟨S,QQᵀ⟩ term; fixed to 25 after grid search.
  • λ (contrastive loss weight) = 5
    Balances LI against LA and LP; fixed to 5.
  • τI, τA (temperatures) = 1
    Contrastive and attention temperatures; both set to 1.
assumptions (3)
  • standard math The first-order Taylor expansion of ⟨S,QQᵀ⟩ around the previous iterate yields a sufficiently accurate surrogate for the non-convex CAOT objective (Appendix A.1).
    Standard successive-linearization technique; convergence is only shown empirically.
  • domain assumption Cosine similarity of predicted probability vectors plus attention similarity of projected embeddings is a faithful proxy for true semantic neighborhood structure among short texts.
    Invoked when S = S_cos + S_att is inserted into Eq. 3; never validated against human judgments.
  • ad hoc to paper A short K-means warm-up on frozen SBERT embeddings supplies pseudo-labels of adequate quality to bootstrap the subsequent EM loop without irreversible error accumulation.
    Stated in Section 3.5 and Algorithm 1; the number of warm-up epochs is dataset-dependent.
invented entities (2)
  • CAOT (consistency-aware adaptive optimal transport) objective
    purpose: Jointly encode sample-to-cluster global structure and sample-to-sample semantic consistency inside a single transport problem whose solution supplies pseudo-labels.
    Defined by Eq. 3; the bilinear consistency term and the soft imbalance regularizer Ψ(b) are the paper’s distinctive modeling choices.
  • Instance-level attention network Gh
    purpose: Produce the learnable similarity matrix S_att that is fed into CAOT and is itself supervised by CAOT pseudo-labels.
    Architecture given in Fig. 3 and Eqs. 5–8; trained with the custom loss LA (Eq. 10).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Beyond Looking Up, Try Looking Around: Harmonizing Global Structure and Local Consistency in Optimal Transport for Short Text Clustering." pith.science (2026). https://pith.science/paper/6BVH7Z75

@misc{pith2026260710548,
  author       = {Pith},
  title        = {Pith review of: Beyond Looking Up, Try Looking Around: Harmonizing Global Structure and Local Consistency in Optimal Transport for Short Text Clustering},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6BVH7Z75}},
  note         = {Machine review of arXiv:2607.10548}
}
read the original abstract

Pseudo-labeling based on Optimal Transport (OT) has become an effective mechanism for enhancing short text clustering. Existing OT methods are short in modeling semantic consistencies between samples, which may assign different pseudo-labels to semantically similar samples. These erroneous pseudo-labels can cause the model to produce inferior clusters. This paper proposes a novel short text clustering framework, which remedies the neglect of semantic consistency in existing OT methods, generating reliable pseudo-labels to facilitate clustering. Specifically, the proposed approach first designs an instance-level attention mechanism to capture semantic relationships between samples, which are then integrated into the OT formulation to endow the transport process with neighborhood semantic awareness. By solving the proposed OT formulation, reliable pseudo-labels are obtained that simultaneously account for sample-to-sample semantic consistency and sample-to-cluster global structure information. These pseudo-labels are then used as supervisory signals to guide the model to achieve accurate clustering. Extensive experiments demonstrate that the proposed approach outperforms state-of-the-art methods. The code is available at: \href{https://github.com/YZH0905/CAOT-STC}{https://github.com/YZH0905/CAOT-STC}.

Figures

Figures reproduced from arXiv: 2607.10548 by the authors.

Figure 1
Figure 1. Schematic illustration of the motivation. Existing methods rely solely on the transport cost from samples to candidate pseudo￾labels (hexagons), where the cost is visualized as the distances from triangles to hexagons, causing semantically similar neighbors to be assigned different pseudo-labels (red boxes). The proposed CAOT addresses this issue by incorporating sample-to-sample semantic consistency, thereby correc… view at source ↗
Figure 2
Figure 2. The overall architecture of the proposed method. It comprises three components: (a) Pseudo-label Generation Module, (b) Semantic Similarity Construction Module, and (c) Supervised Guidance Module. 3.2. The Pseudo-label Generation Module The Pseudo-label Generation Module is expected to produce reliable pseudo-labels. To this end, we propose a novel OT formulation, referred to as CAOT. Solving the optimization proble… view at source ↗
Figure 3
Figure 3. The instance-level attention network. after which the similarity matrix S (1) is computed: S (1) = Softmax  (K (1) 1 K (1)T 2 ) / √ d2  . (6) Finally, the rows of T (1) are linearly combined according to the similarity matrix S (1) as follows: h (1) i = Xn j=1 S (1) ij t (1) j , H(1) = [h (1) 1 ; h (1) 2 ; . . . ;h (1) n ], (7) where t (1) j is the j-th row of T (1), and S (1) ij is the (i, j)-th element of S (1).… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Comparison of representations. Shaded regions indicate the variance derived from 50 runs with different seeds. form poorly due to their inability to produce meaningful representations. (2) Deep neural network-based methods (STCC, Self-Train, SimCSE, ESimCSE, and SCL) o…
Figure 5
Figure 5. Figure 5: The effect of ε1, ε2, ε3, and λ on clustering accuracy. Effectiveness of Pseudo-Label Optimization. We investi￾gate the effectiveness of pseudo-labels for optimizing cluster assignments by removing the LP loss function, i.e., w/o PL. In this variant, the K-means algori…
Figure 6
Figure 6. Figure 6: Convergence validation of the proposed method. The y-axis denotes the change rate of pseudo-labels. B. Supplementary Experiments This section presents extensive supplementary experiments to further validate the effectiveness of the proposed model. §B.1 evaluates the ef…
Figure 7
Figure 7. Figure 7: Visualization of the coupling matrix on StackOverflow. The x-axis represents the sample index, and the y-axis represents the category index. Each column represents the soft prediction of the corresponding sample. On the x-axis, every five consecutive samples belong to …

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

160 extracted references · 5 linked inside Pith

  1. [1]

    Hunter and Kenneth Lange , journal =

    David R. Hunter and Kenneth Lange , journal =. A Tutorial on MM Algorithms , urldate =

  2. [2]

    2004 , publisher=

    Convex optimization , author=. 2004 , publisher=

    2004

  3. [3]

    Natural Hazards , volume=

    Early detection of emergency events from social media: A new text clustering approach , author=. Natural Hazards , volume=. 2022 , publisher=

    2022

  4. [4]

    Applied Sciences , volume=

    Short text clustering algorithms, application and challenges: A survey , author=. Applied Sciences , volume=. 2022 , publisher=

    2022

  5. [5]

    Algorithms , volume=

    Enhancing social media platforms with machine learning algorithms and neural networks , author=. Algorithms , volume=. 2023 , publisher=

    2023

  6. [6]

    Frontiers of Information Technology & Electronic Engineering , volume=

    From Eliza to XiaoIce: challenges and opportunities with social chatbots , author=. Frontiers of Information Technology & Electronic Engineering , volume=. 2018 , publisher=

    2018

  7. [7]

    International Conference on Learning Representations , year=

    Unsupervised clustering using pseudo-semi-supervised learning , author=. International Conference on Learning Representations , year=

  8. [8]

    Advances in neural information processing systems , volume=

    Fixmatch: Simplifying semi-supervised learning with consistency and confidence , author=. Advances in neural information processing systems , volume=

Show all 160 references
  1. [9]

    International Conference on Learning Representations , year=

    ReMixMatch: Semi-Supervised Learning with Distribution Matching and Augmentation Anchoring , author=. International Conference on Learning Representations , year=

  2. [10]

    ArXiv , year=

    FreeMatch: Self-adaptive Thresholding for Semi-supervised Learning , author=. ArXiv , year=

  3. [11]

    Advances in neural information processing systems , volume=

    Unsupervised learning of visual features by contrasting cluster assignments , author=. Advances in neural information processing systems , volume=

  4. [12]

    International conference on machine learning , pages=

    Mixture models for diverse machine translation: Tricks of the trade , author=. International conference on machine learning , pages=. 2019 , organization=

    2019

  5. [13]

    arXiv preprint arXiv:2408.14119 , year=

    Contrastive Learning Subspace for Text Clustering , author=. arXiv preprint arXiv:2408.14119 , year=

    arXiv

  6. [14]

    ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , pages=

    Ot cleaner: Label correction as optimal transport , author=. ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , pages=. 2022 , organization=

    2022

  7. [15]

    Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining , pages=

    A dirichlet multinomial mixture model-based approach for short text clustering , author=. Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining , pages=

  8. [16]

    S im CSE : Simple Contrastive Learning of Sentence Embeddings

    Gao, Tianyu and Yao, Xingcheng and Chen, Danqi. S im CSE : Simple Contrastive Learning of Sentence Embeddings. Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing. 2021

    2021

  9. [17]

    arXiv preprint arXiv:2109.04380 , year=

    Esimcse: Enhanced sample building method for contrastive learning of unsupervised sentence embedding , author=. arXiv preprint arXiv:2109.04380 , year=

    arXiv

  10. [18]

    Advances in Neural Information Processing Systems , volume=

    Csot: Curriculum and structure-aware optimal transport for learning with noisy labels , author=. Advances in Neural Information Processing Systems , volume=

  11. [19]

    2014 , publisher=

    Constrained optimization and Lagrange multiplier methods , author=. 2014 , publisher=

    2014

  12. [20]

    Journal of Machine Learning Research , volume=

    Generalized conditional gradient for sparse estimation , author=. Journal of Machine Learning Research , volume=

  13. [21]

    Foundations and Trends

    Computational optimal transport: With applications to data science , author=. Foundations and Trends. 2019 , publisher=

    2019

  14. [22]

    International Conference on Learning Representations (ICLR) , year=

    Self-labelling via simultaneous clustering and representation learning , author=. International Conference on Learning Representations (ICLR) , year=

  15. [23]

    The Twelfth International Conference on Learning Representations,

    Chuyu Zhang and Hui Ren and Xuming He , title =. The Twelfth International Conference on Learning Representations,

  16. [24]

    Pattern Recognition , volume=

    Exploring uncertainty in pseudo-label guided unsupervised domain adaptation , author=. Pattern Recognition , volume=. 2019 , publisher=

    2019

  17. [25]

    2024 , eprint=

    SP ^2 OT: Semantic-Regularized Progressive Partial Optimal Transport for Imbalanced Clustering , author=. 2024 , eprint=

    2024

  18. [26]

    International conference on machine learning , pages=

    Understanding contrastive representation learning through alignment and uniformity on the hypersphere , author=. International conference on machine learning , pages=. 2020 , organization=

    2020

  19. [27]

    Communication, Simulation, and Intelligent Agents: Implications of Personal Intelligent Machines for Medical Education

    Clancey, William J. Communication, Simulation, and Intelligent Agents: Implications of Personal Intelligent Machines for Medical Education. Proceedings of the Eighth International Joint Conference on Artificial Intelligence (IJCAI-83)

  20. [28]

    Classification Problem Solving

    Clancey, William J. Classification Problem Solving. Proceedings of the Fourth National Conference on Artificial Intelligence

  21. [29]

    , title =

    Robinson, Arthur L. , title =. 1980 , doi =. https://science.sciencemag.org/content/208/4447/1019.full.pdf , journal =

    1980

  22. [30]

    New Ways to Make Microcircuits Smaller---Duplicate Entry

    Robinson, Arthur L. New Ways to Make Microcircuits Smaller---Duplicate Entry. Science

  23. [31]

    Clancey and Glenn Rennels , abstract =

    Diane Warner Hasling and William J. Clancey and Glenn Rennels , abstract =. Strategic explanations for a diagnostic consultation system , journal =. 1984 , issn =. doi:https://doi.org/10.1016/S0020-7373(84)80003-6 , url =

    1984 doi

  24. [32]

    and Rennels, Glenn R

    Hasling, Diane Warner and Clancey, William J. and Rennels, Glenn R. and Test, Thomas. Strategic Explanations in Consultation---Duplicate. The International Journal of Man-Machine Studies

  25. [33]

    Poligon: A System for Parallel Problem Solving

    Rice, James. Poligon: A System for Parallel Problem Solving

  26. [34]

    Transfer of Rule-Based Expertise through a Tutorial Dialogue

    Clancey, William J. Transfer of Rule-Based Expertise through a Tutorial Dialogue

  27. [35]

    The Engineering of Qualitative Models

    Clancey, William J. The Engineering of Qualitative Models

  28. [36]

    Advances in neural information processing systems , volume=

    Attention is all you need , author=. Advances in neural information processing systems , volume=

  29. [37]

    Pluto: The 'Other' Red Planet

    NASA. Pluto: The 'Other' Red Planet

  30. [38]

    IEEE transactions on information theory , volume=

    Least squares quantization in PCM , author=. IEEE transactions on information theory , volume=. 1982 , publisher=

    1982

  31. [39]

    Pattern recognition , volume=

    Gaussian parsimonious clustering models , author=. Pattern recognition , volume=. 1995 , publisher=

    1995

  32. [40]

    International conference on machine learning , pages=

    Unsupervised deep embedding for clustering analysis , author=. International conference on machine learning , pages=. 2016 , organization=

    2016

  33. [41]

    North American Chapter of the Association for Computational Linguistics , year=

    Supporting Clustering with Contrastive Learning , author=. North American Chapter of the Association for Computational Linguistics , year=

  34. [42]

    International Journal of Computer Vision , volume=

    Twin contrastive learning for online clustering , author=. International Journal of Computer Vision , volume=. 2022 , publisher=

    2022

  35. [43]

    International conference on machine learning , pages=

    Barlow twins: Self-supervised learning via redundancy reduction , author=. International conference on machine learning , pages=. 2021 , organization=

    2021

  36. [44]

    International conference on machine learning , pages=

    A simple framework for contrastive learning of visual representations , author=. International conference on machine learning , pages=. 2020 , organization=

    2020

  37. [45]

    Neural Networks , volume=

    Self-taught convolutional neural networks for short text clustering , author=. Neural Networks , volume=. 2017 , publisher=

    2017

  38. [46]

    Advances in Neural Information Processing Systems , volume=

    Leveraging distribution alignment via stein path for cross-domain cold-start recommendation , author=. Advances in Neural Information Processing Systems , volume=

  39. [47]

    arXiv preprint arXiv:2112.09118 , year=

    Unsupervised dense information retrieval with contrastive learning , author=. arXiv preprint arXiv:2112.09118 , year=

    arXiv

  40. [48]

    EDA : Easy Data Augmentation Techniques for Boosting Performance on Text Classification Tasks

    Wei, Jason and Zou, Kai. EDA : Easy Data Augmentation Techniques for Boosting Performance on Text Classification Tasks. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Process...

    2019 doi

  41. [49]

    Journal of big Data , volume=

    Text data augmentation for deep learning , author=. Journal of big Data , volume=. 2021 , publisher=

    2021

  42. [50]

    Proceedings of the European conference on computer vision (ECCV) , pages=

    Deep clustering for unsupervised learning of visual features , author=. Proceedings of the European conference on computer vision (ECCV) , pages=

  43. [51]

    International Journal of Cognitive Computing in Engineering , volume=

    Text clustering with large language model embeddings , author=. International Journal of Cognitive Computing in Engineering , volume=. 2025 , publisher=

    2025

  44. [52]

    Data , volume=

    MN-DS: A multilabeled news dataset for news articles hierarchical classification , author=. Data , volume=. 2023 , publisher=

    2023

  45. [53]

    Lewis, David , title =

  46. [54]

    Pazzani, Michael , title =

  47. [55]

    Mitchell, Tom , title =

  48. [56]

    2009 , publisher=

    Learning multiple layers of features from tiny images , author=. 2009 , publisher=

    2009

  49. [57]

    Proceedings of the fourteenth international conference on artificial intelligence and statistics , pages=

    An analysis of single-layer networks in unsupervised feature learning , author=. Proceedings of the fourteenth international conference on artificial intelligence and statistics , pages=. 2011 , organization=

    2011

  50. [58]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    Class-aware contrastive semi-supervised learning , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

  51. [59]

    Usage of WordNet in natural language processing systems , year=

    Text classification using WordNet hypernyms , author=. Usage of WordNet in natural language processing systems , year=

  52. [60]

    Advances in neural information processing systems , volume=

    Distributed representations of words and phrases and their compositionality , author=. Advances in neural information processing systems , volume=

  53. [61]

    Proceedings of the 4th Workshop on Representation Learning for NLP (RepL4NLP-2019) , pages=

    A self-training approach for short text clustering , author=. Proceedings of the 4th Workshop on Representation Learning for NLP (RepL4NLP-2019) , pages=

    2019

  54. [62]

    Information Sciences , volume=

    Learning deep discriminative representations with pseudo supervision for image clustering , author=. Information Sciences , volume=. 2021 , publisher=

    2021

  55. [63]

    arXiv preprint arXiv:2403.05768 , year=

    Deep Contrastive Multi-view Clustering under Semantic Feature Guidance , author=. arXiv preprint arXiv:2403.05768 , year=

    arXiv

  56. [64]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    Improving unsupervised image clustering with robust learning , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

  57. [65]

    Proceedings of the AAAI conference on artificial intelligence , volume=

    Contrastive clustering , author=. Proceedings of the AAAI conference on artificial intelligence , volume=

  58. [66]

    IEEE Transactions on Image Processing , volume=

    Spice: Semantic pseudo-labeling for image clustering , author=. IEEE Transactions on Image Processing , volume=. 2022 , publisher=

    2022

  59. [67]

    Automatic differentiation in pytorch , author=

  60. [68]

    Natural Language Processing and Information Systems: 25th International Conference on Applications of Natural Language to Information Systems, NLDB 2020, Saarbr

    Enhancement of short text clustering by iterative classification , author=. Natural Language Processing and Information Systems: 25th International Conference on Applications of Natural Language to Information Systems, NLDB 2020, Saarbr. 2020 , organization=

    2020

  61. [69]

    Annual Meeting of the Association for Computational Linguistics , year=

    Robust Representation Learning with Reliable Pseudo-labels Generation via Self-Adaptive Optimal Transport for Short Text Clustering , author=. Annual Meeting of the Association for Computational Linguistics , year=

  62. [70]

    2010 , publisher=

    Introduction to modern information retrieval , author=. 2010 , publisher=

    2010

  63. [71]

    Advances in neural information processing systems , volume=

    Character-level convolutional networks for text classification , author=. Advances in neural information processing systems , volume=

  64. [72]

    Proceedings of the 17th international conference on World Wide Web , pages=

    Learning to classify short and sparse text & web with hidden topics from large-scale data collections , author=. Proceedings of the 17th international conference on World Wide Web , pages=

  65. [73]

    2016 IEEE 32nd International Conference on Data Engineering (ICDE) , pages=

    A model-based approach for text clustering with outlier detection , author=. 2016 IEEE 32nd International Conference on Data Engineering (ICDE) , pages=. 2016 , organization=

    2016

  66. [74]

    1998 , publisher=

    Combinatorial optimization: algorithms and complexity , author=. 1998 , publisher=

    1998

  67. [75]

    Sentence- BERT : Sentence Embeddings using S iamese BERT -Networks

    Reimers, Nils and Gurevych, Iryna. Sentence- BERT : Sentence Embeddings using S iamese BERT -Networks. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019

    2019

  68. [76]

    An RIP Condition for Exact Support Recovery With Covariance-Assisted Matching Pursuit , year=

    Ge, Huanmin and Wang, Libo and Wen, Jinming and Xian, Jun , journal=. An RIP Condition for Exact Support Recovery With Covariance-Assisted Matching Pursuit , year=

  69. [77]

    A Sharp Condition for Exact Support Recovery With Orthogonal Matching Pursuit , year=

    Wen, Jinming and Zhou, Zhengchun and Wang, Jian and Tang, Xiaohu and Mo, Qun , journal=. A Sharp Condition for Exact Support Recovery With Orthogonal Matching Pursuit , year=

  70. [78]

    , journal=

    Tropp, J.A. , journal=. Greed is good: algorithmic results for sparse approximation , year=

  71. [79]

    Multipath Matching Pursuit , year=

    Kwon, Suhyuk and Wang, Jian and Shim, Byonghyo , journal=. Multipath Matching Pursuit , year=

  72. [80]

    Generalized Orthogonal Matching Pursuit , year=

    Wang, Jian and Kwon, Seokbeop and Shim, Byonghyo , journal=. Generalized Orthogonal Matching Pursuit , year=

  73. [81]

    , journal=

    Donoho, D.L. , journal=. Compressed sensing , year=

  74. [82]

    Communications of the ACM , volume=

    Exact matrix completion via convex optimization , author=. Communications of the ACM , volume=. 2012 , publisher=

    2012

  75. [83]

    and Romberg, J

    Candes, E.J. and Romberg, J. and Tao, T. , journal=. Robust uncertainty principles: exact signal reconstruction from highly incomplete frequency information , year=

  76. [84]

    Conference Etrasct 14 Proceedings at International Journal of Engineering Research & Technology , pages=

    Application of DCT in image processing , author=. Conference Etrasct 14 Proceedings at International Journal of Engineering Research & Technology , pages=

  77. [85]

    1999 , publisher=

    A wavelet tour of signal processing , author=. 1999 , publisher=

    1999

  78. [86]

    European Conference on Computer Vision , pages=

    Efficient highly over-complete sparse coding using a mixture model , author=. European Conference on Computer Vision , pages=. 2010 , organization=

    2010

  79. [87]

    IEEE transactions on information theory , volume=

    Decoding by linear programming , author=. IEEE transactions on information theory , volume=. 2005 , publisher=

    2005

  80. [88]

    Comptes rendus mathematique , volume=

    The restricted isometry property and its implications for compressed sensing , author=. Comptes rendus mathematique , volume=. 2008 , publisher=

    2008

  81. [89]

    IEEE Transactions on Signal Processing , title=

    M. IEEE Transactions on Signal Processing , title=. 2006 , volume=

    2006

  82. [90]

    IEEE Transactions on Image Processing , volume=

    Deep k-svd denoising , author=. IEEE Transactions on Image Processing , volume=. 2021 , publisher=

    2021

  83. [91]

    Random Multiplexing for an MIMO-OFDM Radar With Compressed Sensing-Based Reconstruction , year=

    Knill, Christina and Roos, Fabian and Schweizer, Benedikt and Schindler, Daniel and Waldschmidt, Christian , journal=. Random Multiplexing for an MIMO-OFDM Radar With Compressed Sensing-Based Reconstruction , year=

  84. [92]

    and Santos, Juan M

    Lustig, Michael and Donoho, David L. and Santos, Juan M. and Pauly, John M. , journal=. Compressed Sensing MRI , year=

  85. [93]

    Deep residual learning for compressed sensing MRI , year=

    Lee, Dongwook and Yoo, Jaejun and Ye, Jong Chul , booktitle=. Deep residual learning for compressed sensing MRI , year=

  86. [94]

    DAGAN: Deep De-Aliasing Generative Adversarial Networks for Fast Compressed Sensing MRI Reconstruction , year=

    Yang, Guang and Yu, Simiao and Dong, Hao and Slabaugh, Greg and Dragotti, Pier Luigi and Ye, Xujiong and Liu, Fangde and Arridge, Simon and Keegan, Jennifer and Guo, Yike and Firmin, David , journal=. DAGAN: Deep De-Aliasing Generative Adversarial Networks for Fast Compressed ...

  87. [95]

    and Zhang, Yin , journal=

    Li, Chengbo and Sun, Ting and Kelly, Kevin F. and Zhang, Yin , journal=. A Compressive Sensing and Unmixing Scheme for Hyperspectral Data Processing , year=

  88. [96]

    and Natesan Ramamurthy, Karthikeyan and Spanias, Andreas , journal=

    Thiagarajan, Jayaraman J. and Natesan Ramamurthy, Karthikeyan and Spanias, Andreas , journal=. Learning Stable Multilevel Dictionaries for Sparse Representations , year=

  89. [97]

    Direct-Optimization-Based DC Dictionary Learning With the MCP Regularizer , year=

    Li, Zhenni and Yang, Zuyuan and Zhao, Haoli and Xie, Shengli , journal=. Direct-Optimization-Based DC Dictionary Learning With the MCP Regularizer , year=

  90. [98]

    Joint and Direct Optimization for Dictionary Learning in Convolutional Sparse Representation , year=

    Peng, Guan-Ju , journal=. Joint and Direct Optimization for Dictionary Learning in Convolutional Sparse Representation , year=

  91. [99]

    ICML , year=

    Proximal methods for sparse hierarchical dictionary learning , author=. ICML , year=

  92. [100]

    Proceedings of 27th Asilomar conference on signals, systems and computers , pages=

    Orthogonal matching pursuit: Recursive function approximation with applications to wavelet decomposition , author=. Proceedings of 27th Asilomar conference on signals, systems and computers , pages=. 1993 , organization=

    1993

  93. [101]

    Some New Results About Sufficient Conditions for Exact Support Recovery of Sparse Signals via Orthogonal Matching Pursuit , year=

    Liu, Chang and Fang, Yong and Liu, Jianzhong , journal=. Some New Results About Sufficient Conditions for Exact Support Recovery of Sparse Signals via Orthogonal Matching Pursuit , year=

  94. [102]

    Average Performance of Orthogonal Matching Pursuit (OMP) for Sparse Approximation , year=

    Schnass, Karin , journal=. Average Performance of Orthogonal Matching Pursuit (OMP) for Sparse Approximation , year=

  95. [103]

    Signal-Dependent Performance Analysis of Orthogonal Matching Pursuit for Exact Sparse Recovery , year=

    Wen, Jinming and Zhang, Rui and Yu, Wei , journal=. Signal-Dependent Performance Analysis of Orthogonal Matching Pursuit for Exact Sparse Recovery , year=

  96. [104]

    Average Performance of OMP and Thresholding Under Dictionary Mismatch , year=

    Pali, Marie-Christine and Ruetz, Simon and Schnass, Karin , journal=. Average Performance of OMP and Thresholding Under Dictionary Mismatch , year=

  97. [105]

    IEEE Journal of selected topics in signal processing , volume=

    Signal recovery from incomplete and inaccurate measurements via regularized orthogonal matching pursuit , author=. IEEE Journal of selected topics in signal processing , volume=. 2010 , publisher=

    2010

  98. [106]

    and Tsaig, Yaakov and Drori, Iddo and Starck, Jean-Luc , journal=

    Donoho, David L. and Tsaig, Yaakov and Drori, Iddo and Starck, Jean-Luc , journal=. Sparse Solution of Underdetermined Systems of Linear Equations by Stagewise Orthogonal Matching Pursuit , year=

  99. [107]

    Applied and computational harmonic analysis , volume=

    CoSaMP: Iterative signal recovery from incomplete and inaccurate samples , author=. Applied and computational harmonic analysis , volume=. 2009 , publisher=

    2009

  100. [108]

    Signal Recovery from Random Measurements via Extended Orthogonal Matching Pursuit , year=

    Sahoo, Sujit Kumar and Makur, Anamitra , journal=. Signal Recovery from Random Measurements via Extended Orthogonal Matching Pursuit , year=

  101. [109]

    Dictionary Preconditioning for Greedy Algorithms , year=

    Schnass, Karin and Vandergheynst, Pierre , journal=. Dictionary Preconditioning for Greedy Algorithms , year=

  102. [110]

    International Journal of Physical Sciences , volume=

    A re-weighted algorithm for designing data dependent sensing dictionary , author=. International Journal of Physical Sciences , volume=. 2011 , publisher=

    2011

  103. [111]

    IEEE Signal Processing Letters , volume=

    Dictionaries construction using alternating projection method in compressive sensing , author=. IEEE Signal Processing Letters , volume=. 2011 , publisher=

    2011

  104. [112]

    Exact Sparse Signal Recovery via Orthogonal Matching Pursuit with Prior Information , year=

    Wen, Jinming and Yu, Wei , booktitle=. Exact Sparse Signal Recovery via Orthogonal Matching Pursuit with Prior Information , year=

  105. [113]

    Covariance-Assisted Matching Pursuit , year=

    Adler, Amir , journal=. Covariance-Assisted Matching Pursuit , year=

  106. [114]

    An RIP-Based Performance Guarantee of Covariance-Assisted Matching Pursuit , year=

    Wang, Jiayang and Li, Gen and Rencker, Lucas and Wang, Wenwu and Gu, Yuantao , journal=. An RIP-Based Performance Guarantee of Covariance-Assisted Matching Pursuit , year=

  107. [115]

    Sparse Coding From a Bayesian Perspective , year=

    Lu, Xiaoqiang and Wang, Yulong and Yuan, Yuan , journal=. Sparse Coding From a Bayesian Perspective , year=

  108. [116]

    SIAM review , volume=

    Atomic decomposition by basis pursuit , author=. SIAM review , volume=. 2001 , publisher=

    2001

  109. [117]

    and Zhifeng Zhang , journal=

    Mallat, S.G. and Zhifeng Zhang , journal=. Matching pursuits with time-frequency dictionaries , year=

  110. [118]

    and Rao, B.D

    Wipf, D.P. and Rao, B.D. , journal=. Sparse Bayesian learning for basis selection , year=

  111. [119]

    and Wakin, Michael B

    Davenport, Mark A. and Wakin, Michael B. , journal=. Analysis of Orthogonal Matching Pursuit Using the Restricted Isometry Property , year=

  112. [120]

    IEEE Signal Processing Letters , title=

    B. IEEE Signal Processing Letters , title=. 2011 , volume=

    2011

  113. [121]

    International Journal of the Physical Sciences , title=

    A. International Journal of the Physical Sciences , title=. 2011 , volume=

    2011

  114. [122]

    IEEE Signal Processing Letters , title=

    A. IEEE Signal Processing Letters , title=. 2016 , volume=

    2016

  115. [123]

    IEEE Signal Processing Letters , title=

    J. IEEE Signal Processing Letters , title=. 2018 , volume=

    2018

  116. [124]

    Signal Processing , title=

    B. Signal Processing , title=. 2017 , volume=

    2017

  117. [125]

    S. Kay. Fundamentals of Statistical Signal Processing: Estimation Theory. 1993

    1993

  118. [126]

    , journal=

    Baraniuk, Richard G. , journal=. Compressive Sensing [Lecture Notes] , year=

  119. [127]

    and Gilbert, Anna C

    Tropp, Joel A. and Gilbert, Anna C. , journal=. Signal Recovery From Random Measurements Via Orthogonal Matching Pursuit , year=

  120. [128]

    Applied and Computational Harmonic Analysis , volume=

    Iterative thresholding algorithms , author=. Applied and Computational Harmonic Analysis , volume=. 2008 , publisher=

    2008

  121. [129]

    SIAM Journal on numerical analysis , volume=

    Hard thresholding pursuit: an algorithm for compressive sensing , author=. SIAM Journal on numerical analysis , volume=. 2011 , publisher=

    2011

  122. [130]

    Constructive Approximation , volume=

    A simple proof of the restricted isometry property for random matrices , author=. Constructive Approximation , volume=. 2008 , publisher=

    2008

  123. [131]

    Signal Processing , volume=

    A gradient-based alternating minimization approach for optimization of the measurement matrix in compressive sensing , author=. Signal Processing , volume=. 2012 , publisher=

    2012

  124. [132]

    Sensors , volume=

    Comparison of Common Algorithms for Single-Pixel Imaging via Compressed Sensing , author=. Sensors , volume=. 2023 , publisher=

    2023

  125. [133]

    Compressive Radar Imaging , year=

    Baraniuk, Richard and Steeghs, Philippe , booktitle=. Compressive Radar Imaging , year=

  126. [134]

    Signal Processing , volume=

    Sufficient conditions for generalized orthogonal matching pursuit in noisy case , author=. Signal Processing , volume=. 2015 , publisher=

    2015

  127. [135]

    Electronics letters , volume=

    Analysis of generalised orthogonal matching pursuit using restricted isometry constant , author=. Electronics letters , volume=. 2014 , publisher=

    2014

  128. [136]

    International conference on machine learning , pages=

    Co-clustering through optimal transport , author=. International conference on machine learning , pages=. 2017 , organization=

    2017

  129. [137]

    Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

    Cot: Unsupervised domain adaptation with clustering and optimal transport , author=. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

  130. [138]

    Statistics and Computing , volume=

    Robust clustering tools based on optimal transportation , author=. Statistics and Computing , volume=. 2019 , publisher=

    2019

  131. [139]

    Neural Networks , pages=

    Discriminative representation learning via attention-enhanced contrastive learning for short text clustering , author=. Neural Networks , pages=. 2025 , publisher=

    2025

  132. [140]

    Proceedings of the 29th International Conference on Computational Linguistics , pages=

    Esimcse: Enhanced sample building method for contrastive learning of unsupervised sentence embedding , author=. Proceedings of the 29th International Conference on Computational Linguistics , pages=

  133. [141]

    Findings of the Association for Computational Linguistics: EMNLP 2025 , pages=

    FNSCC: Fuzzy Neighborhood-Aware Self-Supervised Contrastive Clustering for Short Text , author=. Findings of the Association for Computational Linguistics: EMNLP 2025 , pages=

    2025

  134. [142]

    2023 , eprint=

    C-Pack: Packaged Resources To Advance General Chinese Embedding , author=. 2023 , eprint=

    2023

  135. [143]

    Improving Transformers with Dynamically Composable Multi-Head Attention , booktitle =

    Da Xiao and Qingye Meng and Shengping Li and Xingyuan Yuan , editor =. Improving Transformers with Dynamically Composable Multi-Head Attention , booktitle =. 2024 , url =

    2024

  136. [144]

    arXiv preprint arXiv:2506.05176 , year=

    Qwen3 embedding: Advancing text embedding and reranking through foundation models , author=. arXiv preprint arXiv:2506.05176 , year=

    arXiv

  137. [145]

    Engineering Applications of Artificial Intelligence , volume=

    Constructing a robust Short-Text Clustering Model for contrastive learning based on optimized adaptive optimal transport for pseudo-label generation , author=. Engineering Applications of Artificial Intelligence , volume=. 2025 , publisher=

    2025

  138. [146]

    1997 , howpublished =

    Mitchell, Tom , title =. 1997 , howpublished =

    1997

  139. [147]

    Quality & Quantity , pages=

    A Hybrid Model Integrating LDA, BERT, and Clustering for Enhanced Topic Modeling , author=. Quality & Quantity , pages=. 2025 , publisher=

    2025

  140. [148]

    ACM computing surveys (CSUR) , volume=

    Data clustering: a review , author=. ACM computing surveys (CSUR) , volume=. 1999 , publisher=

    1999

  141. [149]

    Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2 (Short Papers) , pages=

    Contextual augmentation: Data augmentation by words with paradigmatic relations , author=. Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2 (Short Papers) , pages=

    2018

  142. [150]

    Pattern Recognition , pages=

    discdc: Unsupervised discriminative deep image clustering via confidence-driven self-labeling , author=. Pattern Recognition , pages=. 2025 , publisher=

    2025

  143. [151]

    Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track) , pages=

    A large-scale real-world evaluation of an llm-based virtual teaching assistant , author=. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track) , pages=

  144. [152]

    2026 , url=

    Jiayu Qin and Zhengquan Luo and Guy Tadmor and Changyou Chen and David Zeevi and zhiqiang xu , booktitle=. 2026 , url=

    2026

  145. [153]

    IEEE Transactions on Circuits and Systems for Video Technology , year=

    Adaptive Pseudo-label Purification and Debiasing for Unsupervised Visible-Infrared Person Re-Identification , author=. IEEE Transactions on Circuits and Systems for Video Technology , year=

  146. [154]

    Proceedings of the AAAI Conference on Artificial Intelligence , volume=

    Multi-level optimal transport for universal cross-tokenizer knowledge distillation on language models , author=. Proceedings of the AAAI Conference on Artificial Intelligence , volume=

  147. [155]

    Journal of the royal statistical society: series B (methodological) , volume=

    Maximum likelihood from incomplete data via the EM algorithm , author=. Journal of the royal statistical society: series B (methodological) , volume=. 1977 , publisher=

    1977

  148. [156]

    IEEE Transactions on Pattern Analysis and Machine Intelligence , year=

    ClusMatch: Improving deep clustering by unified positive and negative pseudo-label learning , author=. IEEE Transactions on Pattern Analysis and Machine Intelligence , year=

  149. [157]

    International Journal of Human--Computer Interaction , pages=

    User Needs and Benefits of Mental Health Chatbots: Text-Mining Analysis of Mobile Apps Reviews , author=. International Journal of Human--Computer Interaction , pages=. 2025 , publisher=

    2025

  150. [158]

    Journal of Food Protection , volume=

    Exploring food safety emergency incidents on Sina Weibo: using text mining and sentiment evolution , author=. Journal of Food Protection , volume=. 2025 , publisher=

    2025

  151. [159]

    IEEE Transactions on Neural Networks and Learning Systems , year=

    Improving image contrastive clustering through self-learning pairwise constraints , author=. IEEE Transactions on Neural Networks and Learning Systems , year=

  152. [160]

    IEEE Transactions on Pattern Analysis and Machine Intelligence , year=

    ClusMatch: Improving Deep Clustering by Unified Positive and Negative Pseudo-label Learning , author=. IEEE Transactions on Pattern Analysis and Machine Intelligence , year=

Pith tools

Reviewed July 14, 2026 · model on record in the stance chip above.