Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

Benchmarking Graph Neural Networks for Document Layout Analysis in Public Affairs

T0 review · 4 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read GraphSAGE with local links leads GNN benchmark for official document layout

desk verdict A useful internal benchmark of GNN variants for Spanish official-document layout analysis, but the headline comparison to the Random Forest baseline is not supported by the evidence as presented. read the letter →

arxiv 2505.14699 v2 pith:WXFF3NAT submitted 2025-05-12 cs.CV cs.CLcs.LG

Miguel Lopez-Duran , Julian Fierrez , Aythami Morales , Ruben Tolosana , Oscar Delgado-Mohatar , Alvaro Ortigosa This is my paper · ORCID
classification cs.CVcs.CLcs.LG
keywords GraphNeuralNetworksDocumentLayoutAnalysisdigital-bornPDFmultimodalfusionSAGEk-closestneighborpublicaffairsdocumentstextblockclassification
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

The paper asks whether graph neural networks can classify the semantic role of text blocks in born-digital official documents, such as Spanish gazettes, where layout objects are exactly placed and need no OCR. It claims that among the tested GNN configurations, GraphSAGE operating on a k-closest-neighbor graph with text and visual features processed in separate branches achieves the highest per-class and overall accuracy, and that it beats a Random Forest baseline on some sources and on overall accuracy in about one third of them. If true, this would show that local layout relationships and multimodal fusion are exploitable by GNNs for fine-grained layout analysis, without manual feature engineering.

What carries the argument

The central object is the page-level document graph, where each layout object is a node and edges are built either by connecting each node to its k closest neighbors by centroid distance (k=4 in all experiments) or by connecting all nodes. Node features come from two pretrained models: a Spanish RoBERTa-base for text and a ResNet-18 for cropped page regions. The GNN architectures—GCN, GAT, GraphSAGE, and TAGCN—are then run in three frameworks (text-only, visual-only, concatenated, and dual-branch), with the dual-branch framework using separate text and visual GNNs whose outputs are fused before classification.

What would settle it

Retrain the Random Forest baseline on the same 80/20 validation split used for the GNNs and compare per-source overall accuracy; if the baseline then matches or exceeds GraphSAGE on these sources, the claimed gains over the baseline would not hold under a matched protocol.

Watch

Extended reading notes

Core claim

The paper's central claim is that a dual-branch GraphSAGE model on a k-closest-neighbor document graph performs best among the GNN settings tested, and that this configuration improves per-class accuracy over the Random Forest baseline for some classes in most sources and overall accuracy in one third of the sources. The discovery is framed as the first GNN-based layout analysis applied to Spanish official digital-born documents, with the finding that local neighborhood structure and separate text and visual processing branches are the key ingredients for the gains.

Load-bearing premise

The protocol assumes the PAL validation set is reliable and sufficient for training the GNNs, while the original training set used for the compared baseline is too unreliable to use.

Editorial extensions

If this is right

  • If local layout relationships are as informative as the paper suggests, then graph-based methods can be used instead of manual layout heuristics for born-digital documents, with node features from pretrained transformers.
  • The dual-branch fusion pattern—separate text and visual GNNs followed by concatenation—could transfer to other document types and languages where the same four semantic classes apply.
  • The k-closest graph with small k=4 keeps the graph diameter small while focusing on local context, implying that only nearby layout objects matter most for classifying a text block.
  • The benchmark provides a fixed evaluation protocol on a multi-source public affairs dataset, which could serve as a comparison point for future layout analysis methods on digital-born official documents.
  • The finding that GraphSAGE outperforms GCN, GAT, and TAGCN in these settings suggests that inductive neighborhood aggregation is well suited to layout graphs with variable structure across pages and sources.

Reading between the lines

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

  • Because Table 1 shows the Random Forest baseline averages 97.48% overall while the best GNN averages 95.98%, the claim of beating the baseline rests entirely on source-level and per-class comparisons; a matched Random Forest trained on the same validation split could well erase or reverse the reported gains.
  • The choice k=4 is not tested for sensitivity; probing k values from 2 to 8 on a few sources would indicate whether the result is robust or tied to that specific neighborhood size.
  • The paper's decision to train on the validation set while citing the original training set as unreliable implicitly questions the baseline's training source; a fairer comparison would retrain the baseline under the same protocol.
  • The method could plausibly extend to classify non-text objects (images, tables, links) rather than only text blocks, since those nodes are retained in the graph and their embeddings are updated during training.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper benchmarks four GNN architectures (GCN, GAT, GraphSAGE, TAGCN) for fine-grained classification of text-block layout categories (Identifier, Title, Summary, Body) in digital-born official documents from the PAL dataset. It introduces two page-level graph constructions (k-closest-neighbor and complete), initializes node features with pretrained RoBERTa (text) and ResNet (vision) embeddings, and evaluates three frameworks: text-only, vision-only, concatenated single-branch, and dual-branch multimodal. Experiments are per-source, with 80/20 document-level splits and 5-fold cross-validation on the PAL validation set. The authors report that a dual-branch GraphSAGE on the k-closest graph achieves the best GNN accuracy and claim it outperforms the Random Forest baseline in some sources and in about one third of sources overall.

Significance. If the central comparison were well controlled, the paper would provide a useful benchmark for graph-based multimodal layout analysis in a domain (public affairs documents) with practical value. The use of pretrained text and vision embeddings avoids manual feature engineering, the k-closest graph is a sensible inductive bias for local layout relations, and the comparison across four GNN backbones and three fusion schemes is informative. The paper also ships a public dataset link and states reproducible experimental choices. However, the significance is materially reduced by the uncontrolled baseline comparison: the GNNs are trained on the validation set while the Random Forest baseline was trained on the original training set, and the per-source comparisons that underlie the headline claims are not shown in any table. As presented, the strongest defensible claim is limited to relative ranking among the tested GNN configurations, not superiority over the baseline.

major comments (4)
  1. [Section 4.2 and Table 1] The comparison against the Random Forest baseline is not protocol-matched. The GNN models are trained and tested exclusively on the PAL validation set (80/20 document split per source, Section 4.2), while the baseline numbers in Table 1 come from models trained on the original training set of PAL, as described in Section 4.1. If the validation set differs in label quality or distribution, any improvement over the baseline could be due to the training-data choice rather than to graph-based modeling. Please retrain the Random Forest under the same validation-set protocol, or clearly restrict all claims to a comparison among GNN variants and add an explicit statement that no controlled comparison with the baseline is claimed.
  2. [Table 2 and Section 4.3] The headline claim that the proposed model 'outperforms the baseline in some sources' is not verifiable from the reported results. Table 2 reports accuracy for the five selected sources but contains no baseline column, and Table 1 reports only source-averaged accuracy, where the Random Forest baseline (97.48% overall) exceeds every GNN configuration (best dual-branch 95.98%). Please provide a per-source table that includes the baseline accuracy for each source and each class, so that the claimed per-source improvements can be checked.
  3. [Section 5 vs. Section 4.3] The conclusion states that the dual-branch GraphSAGE configuration improves overall accuracy 'in one third of them' (Section 5), but the results section states that, among the five selected sources, the model achieves higher overall accuracy 'in one of the cases' (Section 4.3), which is 20%, not one third. These statements are inconsistent, and neither is backed by a per-source comparison table. Please harmonize the quantification and support it with the actual per-source baseline comparison.
  4. [Table 1 and Section 4.3] The claim that the dual-branch framework outperforms the other frameworks in four of the five top configurations is based on selecting the best configuration per framework after seeing the results. The paper acknowledges this selection bias in general terms, but the central comparison to the baseline is still affected by it, since the baseline is a single fixed model. Please report results for all configurations or apply a multiple-comparison correction, and provide error bars or confidence intervals for the aggregated accuracies, given the large variance implied by the per-source variation.
minor comments (4)
  1. [Section 4.1] The abstract and introduction mention 'more than 20 sources' and '20 sources', while Section 4.1 states the dataset contains 24 official gazettes; please make the source count consistent.
  2. [Table 2] The table uses 'Páis Vasco' but the correct spelling is 'País Vasco'; also, the formatting of the mean and standard deviation values (e.g., '99, 520,76') is inconsistent and should be cleaned to avoid confusion.
  3. [Section 3.2] The caption of Figure 2 says 'four different models are trained independently' while the figure illustrates three frameworks; please clarify the caption to avoid ambiguity.
  4. [Section 3.1] The choice of k=4 in the k-closest graph is stated without any sensitivity analysis; a brief comment on how the results vary with k would strengthen the benchmark.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central claim is an empirical benchmark against an external dataset and baseline, with no fitted quantity or self-citation chain doing the work.

full rationale

The paper's derivation chain is an empirical benchmark rather than a formal derivation. Node features come from frozen pretrained external models (RoBERTa-base and ResNet-18), graph structures are defined from geometric proximity, the GNN backbones are standard architectures, and the target labels come from the PAL dataset. No equation defines a prediction in terms of the quantity it is claimed to predict, and no fitted parameter is renamed as a result. The only self-citations are to the PAL dataset and the Random Forest baseline of Peña et al. [26,27], which are used as experimental references and feature-backbone motivation; these are not load-bearing justifications of the GNN comparison, so they do not make the claim circular. The paper itself flags the main threat in Section 4.3: the averaged results "may present an overly optimistic view" because the best configuration is selected per framework. That is a selection-bias limitation, not a fitted-input-called-prediction cycle. A further concern is protocol mismatch: GNNs are trained and tested on the PAL validation set (Section 4.2) while the baseline numbers come from models trained on the PAL training set (Section 4.1), and Table 2 omits baseline columns, so the "outperforms baseline in some sources" claim is difficult to verify from the reported tables. These are correctness and reporting risks, not circularity: the comparison is not equivalent to its own inputs by construction. Accordingly, no circular step is identified and the score is 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new theoretical entities or physical objects. Its central claim rests on domain assumptions about layout semantics, multimodal complementarity, and the quality of the PAL validation set, plus a small number of hand-chosen hyperparameters. The most consequential assumption is that the validation set can substitute for a proper training set, since the baseline comparison depends on it.

free parameters (3)
  • k in k-closest graph = 4
    Set by hand in Section 3.1 to focus on local dependencies while keeping graph diameter small; not tuned per source or cross-validated.
  • Batch size per source = not reported
    Section 4.2 says the batch size is manually adjusted per source to ensure stable training, but the actual values are not provided.
  • Class weights for weighted cross-entropy = not reported
    Weighted cross-entropy is used in Section 4.2, but the weight values or how they were derived are not given.
assumptions (4)
  • domain assumption The semantic category of a layout object is significantly influenced by its neighboring elements on the page.
    Invoked in Section 3.1 to motivate the k-closest graph. If local neighbors carry little information about a block's class, the GNN graph structure adds no value.
  • domain assumption Combining textual and visual features improves layout classification.
    Used throughout Section 3.2 to justify the multimodal frameworks. If one modality is sufficient, the dual-branch and concatenated models are unnecessary.
  • domain assumption The PAL validation set labels are accurate enough for training and testing, while the original PAL training set is not reliable.
    Section 4.1 states the training set may contain labeling inaccuracies and therefore only the validation set is used. The whole comparison against the baseline depends on this claim.
  • domain assumption Digital-born PDFs have exact layout object positions, so no OCR is needed.
    Mentioned in Section 2.3. This distinguishes the task from scanned document analysis and justifies using bounding boxes from PDF metadata.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Benchmarking Graph Neural Networks for Document Layout Analysis in Public Affairs." pith.science (2026). https://pith.science/paper/WXFF3NAT

@misc{pith2026250514699,
  author       = {Pith},
  title        = {Pith review of: Benchmarking Graph Neural Networks for Document Layout Analysis in Public Affairs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WXFF3NAT}},
  note         = {Machine review of arXiv:2505.14699}
}
read the original abstract

The automatic analysis of document layouts in digital-born PDF documents remains a challenging problem due to the heterogeneous arrangement of textual and nontextual elements and the imprecision of the textual metadata in the Portable Document Format. In this work, we benchmark Graph Neural Network (GNN) architectures for the task of fine-grained layout classification of text blocks from digital native documents. We introduce two graph construction structures: a k-closest-neighbor graph and a fully connected graph, and generate node features via pre-trained text and vision models, thus avoiding manual feature engineering. Three experimental frameworks are evaluated: single-modality (text or visual), concatenated multimodal, and dual-branch multimodal. We evaluated four foundational GNN models and compared them with the baseline. Our experiments are specifically conducted on a rich dataset of public affairs documents that includes more than 20 sources (e.g., regional and national-level official gazettes), 37K PDF documents, with 441K pages in total. Our results demonstrate that GraphSAGE operating on the k-closest-neighbor graph in a dual-branch configuration achieves the highest per-class and overall accuracy, outperforming the baseline in some sources. These findings confirm the importance of local layout relationships and multimodal fusion exploited through GNNs for the analysis of native digital document layouts.

Figures

Figures reproduced from arXiv: 2505.14699 by the authors.

Figure 1
Figure 1. Illustration of the graph construction process applied to a document from the Diario Oficial de la Generalitat Valenciana. Original document page (a), k-closest graph (b), and complete graph (c) representation of the page [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Overview of the three experimental frameworks. After passing each object through the transformers, four different models are trained independently to perform three different classifications of the text block object [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Visual examples of page document images from different official gazettes (upper row). From left to right we present: i) Spanish official gazette (i.e., BOE); ii) public affairs document with the results of a vote in the Spanish Parliament; iii) Spanish regional official gazette (i.e. DOGV); and iv) Official Journal of the European Union. The bottom row presents illustrations of their layout components simplified by … view at source ↗

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. Privacy-Aware Detection of Fake Identity Documents: Methodology, Benchmark, and Improved Algorithms (FakeIDet2)

    cs.CR 2025-08 conditional novelty 6.0 of 10

    Patch-based training on anonymized ID fragments detects physical and synthetic fake IDs across unseen datasets without sharing full identity documents.

Reference graph

Works this paper leans on

40 extracted references · 33 canonical work pages · cited by 1 Pith paper

  1. [1]

    In: Proc

    Banerjee, A., Biswas, S., Lladós, J., Pal, U.: GraphKD: Exploring Knowledge Dis- tillation Towards Document Object Detection with Structured Graph Creation. In: Proc. of ICDAR. pp. 354–373. Springer (2024)

    work page 2024

  2. [2]

    In: ACM/IEEE Joint Conf

    Bast, H., Korzen, C.: A benchmark and evaluation for text extraction from PDF. In: ACM/IEEE Joint Conf. on Digital Libraries (JCDL). pp. 1–10 (2017)

    work page 2017

  3. [3]

    Advances in Neural Information Processing Systems33, 1877–1901 (2020)

    Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J.D., Dhariwal, P., Nee- lakantan, A., Shyam, P., Sastry, G., Askell, A., et al.: Language models are few-shot learners. Advances in Neural Information Processing Systems33, 1877–1901 (2020)

    2020

  4. [4]

    In: Proc

    Da, C., Luo, C., Zheng, Q., Yao, C.: Vision grid transformer for document layout analysis. In: Proc. ICCV. pp. 19462–19472 (2023)

    work page 2023

  5. [5]

    In: 2020 25th Inter- national Conference on Pattern Recognition (ICPR)

    Dang, T.A.N., Hoang, D.T., Tran, Q.B., Pan, C.W., Nguyen, T.D.: End-to-end hi- erarchical relation extraction for generic form understanding. In: 2020 25th Inter- national Conference on Pattern Recognition (ICPR). pp. 5238–5245. IEEE (2021) 14 M. Lopez-Duran, J. Fierrez, A. Morales, et al

    work page 2021

  6. [6]

    arXiv preprint arXiv:1710.10370 (2017)

    Du, J., Zhang, S., Wu, G., Moura, J.M., Kar, S.: Topology adaptive graph convo- lutional networks. arXiv preprint arXiv:1710.10370 (2017)

    arXiv 2017

  7. [7]

    Part 2: Trends and challenges

    Fierrez, J., Morales, A., Vera-Rodriguez, R., Camacho, D.: Multiple classifiers in biometrics. Part 2: Trends and challenges. Information Fusion44, 103–112 (2018)

    work page 2018

  8. [8]

    In: European Conference on Computer Vision

    Gemelli, A., Biswas, S., Civitelli, E., Lladós, J., Marinai, S.: Doc2graph: a task agnostic document understanding framework based on graph neural networks. In: European Conference on Computer Vision. pp. 329–344. Springer (2022)

    work page 2022

Show all 40 references
  1. [9]

    In: 2022 26th International Conference on Pattern Recognition (ICPR)

    Gemelli, A., Vivoli, E., Marinai, S.: Graph neural networks and representation embedding for table extraction in PDF documents. In: 2022 26th International Conference on Pattern Recognition (ICPR). pp. 1719–1726. IEEE (2022)

    2022

  2. [10]

    Gu, Z., Meng, C., Wang, K., Lan, J., Wang, W., Gu, M., Zhang, L.: XYLayoutLM: Towardslayout-awaremultimodalnetworksforvisually-richdocumentunderstand- ing.In:ProceedingsoftheIEEE/CVFConferenceonComputerVisionandPattern Recognition. pp. 4583–4592 (2022)

    2022

  3. [11]

    Procesamiento del Lenguaje Natural68 (2022)

    Gutiérrez Fandiño, A., et al.: MarIA: Spanish Language Models. Procesamiento del Lenguaje Natural68 (2022)

    2022

  4. [12]

    Advances in Neural Information Processing Systems30 (2017)

    Hamilton, W., Ying, Z., Leskovec, J.: Inductive representation learning on large graphs. Advances in Neural Information Processing Systems30 (2017)

    2017

  5. [13]

    In: Proc

    He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: Proc. CVPR. pp. 770–778 (2016)

    2016

  6. [14]

    Standard (2008)

    ISO: Document management—-Portable document format—-Part 1: PDF 1.7. Standard (2008)

    2008

  7. [15]

    In: Computer Science & Information Technology (CS & IT)

    Josi, F., Wartena, C., Heid, U.: Preparing legal documents for NLP analysis: Im- proving the classification of text elements by using page features. In: Computer Science & Information Technology (CS & IT). pp. 17–29. AIRCC Publishing Cor- poration (2022)

    2022

  8. [16]

    In: Proceedings of NAACL-HLT

    Kenton, J.D.M.W.C., Toutanova, L.K.: BERT: Pre-training of deep bidirectional transformers for language understanding. In: Proceedings of NAACL-HLT. vol. 1. Minneapolis, Minnesota (2019)

    2019

  9. [17]

    arXiv preprint arXiv:1609.02907 (2016)

    Kipf, T.N., Welling, M.: Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907 (2016)

    2016 arXiv

  10. [18]

    IEEE Access11, 117576–117581 (2023)

    Kutbi, M.: Named entity recognition utilized to enhance text classification while preserving privacy. IEEE Access11, 117576–117581 (2023)

    2023

  11. [19]

    IEEE Signal Processing Magazine 40(4), 49–60 (2023)

    Leus, G., Marques, A.G., Moura, J.M., Ortega, A., Shuman, D.I.: Graph signal processing: History, development, impact, and outlook. IEEE Signal Processing Magazine 40(4), 49–60 (2023)

    2023

  12. [20]

    In: Proceedings of the ACM Web Conference 2022

    Li, S., Gou, G., Liu, C., Hou, C., Li, Z., Xiong, G.: TTAGN: Temporal transaction aggregation graph network for ethereum phishing scams detection. In: Proceedings of the ACM Web Conference 2022. pp. 661–669 (2022)

    2022

  13. [21]

    Neurocomputing549, 126441 (2023)

    Li, X., Sun, L., Ling, M., Peng, Y.: A survey of graph neural network based rec- ommendation in social networks. Neurocomputing549, 126441 (2023)

    2023

  14. [22]

    IEEE Trans

    Liang, G., Tiwari, P., et al.: Dynamic Causal Explanation Based Diffusion- Variational Graph Neural Network for Spatiotemporal Forecasting. IEEE Trans. on Neural Networks and Learning Systems (2024)

    2024

  15. [23]

    arXiv preprint arXiv:1907.11692 (2019)

    Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., Levy, O., Lewis, M., Zettlemoyer, L., Stoyanov, V.: RoBERTa: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692 (2019)

    2019 arXiv

  16. [24]

    arXiv preprint arXiv:2503.07384 (2025) Benchmarking Graph Neural Networks for Document Layout Analysis 15

    Mancera, G., DeAlcala, D., Fierrez, J., Tolosana, R., Morales, A.: Is my text in your AI model? Gradient-based membership inference test applied to LLMs. arXiv preprint arXiv:2503.07384 (2025) Benchmarking Graph Neural Networks for Document Layout Analysis 15

    2025 arXiv

  17. [25]

    In: IAPR Intl

    Mancera, G., Morales, A., Fierrez, J., Tolosana, R., Penna, A., Lopez-Duran, M., Jurado, F., Ortigosa, A.: PBa-LLM: Privacy- and bias-aware NLP using named- entity recognition (NER). In: IAPR Intl. Conf. on Document Analysis and Recog- nition Workshops (ICDARw) (September 2025)

    2025

  18. [26]

    Information Fusion 108, 102398 (2024)

    Peña, A., Morales, A., Fierrez, J., Ortega-Garcia, J., Puente, I., Cordova, J., Cor- dova, G.: Continuous document layout analysis: Human-in-the-loop AI-based data curation, database, and evaluation in the domain of public affairs. Information Fusion 108, 102398 (2024)

    2024

  19. [27]

    In: International Conference on Document Analysis and Recognition

    Peña, A., Morales, A., Fierrez, J., Serna, I., Ortega-Garcia, J., Puente, I., Cordova, J., Cordova, G.: Leveraging large language models for topic classification in the domain of public affairs. In: International Conference on Document Analysis and Recognition. pp. 20–33. Spri...

    2023

  20. [28]

    In: AAAI/ACM Conf

    Peña, A., Fierrez, J., Morales, A., Mancera, G., Lopez, M., Tolosana, R.: Ad- dressing bias in LLMs: Strategies and application to fair AI-based recruitment. In: AAAI/ACM Conf. on AI, Ethics, and Society (AIES) (October 2025)

    2025

  21. [29]

    SN Computer Science4(5), 434 (June 2023)

    Peña,A.,Serna,I.,Morales,A.,Fierrez,J.,Ortega,A.,Herrarte,A.,Alcantara,M., Ortega-Garcia, J.: Human-centric multimodal machine learning: Recent advances and testbed on AI-based recruitment. SN Computer Science4(5), 434 (June 2023)

    2023

  22. [30]

    In: 2019 International Conference on Document Analysis and Recognition (ICDAR)

    Qasim, S.R., Mahmood, H., Shafait, F.: Rethinking table recognition using graph neural networks. In: 2019 International Conference on Document Analysis and Recognition (ICDAR). pp. 142–147. IEEE (2019)

    2019

  23. [31]

    Pattern Recognition 127, 108641 (2022)

    Riba, P., Goldmann, L., Terrades, O.R., Rusticus, D., Fornés, A., Lladós, J.: Ta- ble detection in business document images by message passing networks. Pattern Recognition 127, 108641 (2022)

    2022

  24. [32]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Tang, Z., Yang, Z., Wang, G., Fang, Y., Liu, Y., Zhu, C., Zeng, M., Zhang, C., Bansal, M.: Unifying vision, text, and layout for universal document processing. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 19254–19264 (2023)

    2023

  25. [33]

    In: European Conf

    Tello, J., de la Cruz, M., Ribeiro, T., Fierrez, J., Morales, A., Tolosana, R., Alonso, C.L., Ortega, A.: Symbolic AI (LFIT) for XAI to handle biases. In: European Conf. on AI Workshops (ECAIw). CEUR-WS, vol. 3523 (2023)

    2023

  26. [34]

    In: ICLR (2018)

    Veličković, P., Cucurull, G., Casanova, A., Romero, A., Liò, P., Bengio, Y.: Graph Attention Networks. In: ICLR (2018)

    2018

  27. [35]

    Artificial Intelligence Review 56(7), 6295–6364 (2023)

    Waikhom, L., Patgiri, R.: A survey of graph neural networks in various learning paradigms: methods, applications, and challenges. Artificial Intelligence Review 56(7), 6295–6364 (2023)

    2023

  28. [36]

    Foundations and Trends® in Machine Learning16(2), 119–328 (2023)

    Wu, L., Chen, Y., Shen, K., Guo, X., Gao, H., Li, S., Pei, J., Long, B., et al.: Graph neural networks for natural language processing: A survey. Foundations and Trends® in Machine Learning16(2), 119–328 (2023)

    2023

  29. [37]

    IEEE Journal of Biomedical and Health Informatics 28(4), 1836–1847 (2023)

    Xie, Q., Tiwari, P., Ananiadou, S.: Knowledge-enhanced graph topic transformer for explainable biomedical text summarization. IEEE Journal of Biomedical and Health Informatics 28(4), 1836–1847 (2023)

    2023

  30. [38]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Zhang, N., Cheng, H., Chen, J., Jiang, Z., Huang, J., Xue, Y., Jin, L.: M2Doc: a multi-modal fusion approach for document layout analysis. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 38, pp. 7233–7241 (2024)

    2024

  31. [39]

    In: Proc

    Zhang, Y., Yu, X., Cui, Z., Wu, S., Wen, Z., Wang, L.: Every Document Owns Its Structure: Inductive Text Classification via Graph Neural Networks. In: Proc. ACL. pp. 334–339 (2020)

    2020

  32. [40]

    In: 2019 International Conference on Document Analysis and Recog- nition (ICDAR)

    Zhong,X.,Tang,J.,Yepes,A.J.:PubLayNet:largestdataseteverfordocumentlay- out analysis. In: 2019 International Conference on Document Analysis and Recog- nition (ICDAR). pp. 1015–1022. IEEE (2019)

    2019

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.