Pith. sign in

REVIEW 6 major objections 5 minor 53 references

Your Next State-of-the-Art Could Come from Another Domain: A Cross-Domain Analysis of Hierarchical Text Classification

T0 review · 6 major / 5 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 State-of-the-art hierarchical text classifiers often come from other domains, this paper argues, and dataset traits matter more than domain origin.

desk verdict Useful cross-domain benchmark and framework; the new-SOTA claims are pipeline-relative and provisional, but the cross-domain transfer insight is solid. read the letter →

arxiv 2412.12744 v1 pith:5E2QLUQ6 submitted 2024-12-17 cs.CL cs.AIcs.LG

Nan Li , Bo Kang , Tijl De Bie This is my paper · ORCID
classification cs.CLcs.AIcs.LG
keywords hierarchicaltextclassificationmulti-labelcross-domainanalysisunifiedevaluationframeworkextremedatasetcharacteristicslabelhierarchystate-of-the-artcomparison
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 argues that for hierarchical text classification—assigning labels from a tree-structured vocabulary, as in medical coding, patent tagging, or legal indexing—the best-performing method on a given dataset is often built for a different domain than the data. It backs this with a unified evaluation of eight state-of-the-art methods across eight datasets from law, science, news, medicine, and patents, all processed through one shared pipeline. Under that pipeline, methods from the legal and extreme multi-label domains, XR-Transformer and MatchXML, surpass domain-native leaders on news, science, and patent benchmarks, and a medical method, PLM-ICD, ranks near the top on legal data. The paper concludes that dataset characteristics such as label count, document length, and training-set size matter more than the domain a method was designed for, and that recombining submodules across domains can yield new state of the art.

What carries the argument

The argument is carried by a unified framework that decomposes every hierarchical text classification method into nine submodules: data augmentation, long-document handling, text encoder, label encoder, text-label information fusion, output-space segmentation, training objectives, score prediction, and prediction refinement. Each of the 32 surveyed methods is positioned by its specific choices within these submodules, allowing methods from different domains to be compared on equal architectural terms. The framework also distinguishes explicit use of a given taxonomy from implicit label-structure learning, which is what makes cross-domain submodule recombination—such as grafting MatchXML's label semantics onto PLM-ICD's document handling—a concrete, testable operation.

What would settle it

Re-run the eight methods with unrestricted per-method hyperparameter search and the original, unmodified dataset hierarchies; if domain-native methods such as HILL on NYT-166 and THMM on USPTO2M-632 reclaim the top ranks, the reported cross-domain SOTA would be an artifact of the unified pipeline's constraints rather than a property of the methods.

Watch

Extended reading notes

Core claim

The central claim is that state-of-the-art hierarchical text classification performance is not owned by any single domain: the paper reports new state-of-the-art results on NYT-166, SciHTC-83, USPTO2M-632, and MIMIC3-3681 by applying methods from other domains or by combining submodules across domains. In the unified evaluation, XR-Transformer and MatchXML, both originally built for extreme multi-label legal text, overtake the previous news, science, and patent leaders, while PLM-ICD, a medical-coding model, becomes a top-three performer on legal and news data. The paper further shows that adding Label2Vec from MatchXML to PLM-ICD produces a new MIMIC3-3681 state of the art, and that a medical-domain language model improves patent classification more than a scientific-domain model does. The explanation the paper offers is that dataset characteristics—label-space size, label cardinality, document length, training-data volume, and label-pattern diversity—predict method effectiveness better than the domain of origin.

Load-bearing premise

The unified evaluation pipeline—modified label hierarchies on MIMIC3-3681 and USPTO2M-632, limited hyperparameter search, and 36-hour/40GB resource caps—does not systematically handicap domain-native methods.

Editorial extensions

If this is right

  • Researchers should benchmark candidate methods from other domains before committing to a domain-specific architecture, since cross-domain methods set the strongest results on four of the eight datasets.
  • Dataset traits such as label count, average document length, and training-set size can be used to choose which family of methods is likely to work best, rather than defaulting to the method developed for the data's own domain.
  • Combining submodules from different domains, such as PLM-ICD's long-document segmentation with MatchXML's Label2Vec label semantics, can produce a new state of the art and is a practical route to better systems.
  • Domain-specific language models are valuable, especially for simple architectures and challenging datasets, but their benefit can transfer across domains in unexpected ways, as when a medical model outperforms a scientific model on patents.
  • Long-document handling requires careful design: simple truncation hurts medical coding, while naive averaging over segments also hurts, suggesting that effective strategies must be sophisticated and label-aware.

Reading between the lines

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

  • The observed transfer pattern suggests that common label-space geometry—label count, co-occurrence sparsity, and label-pattern diversity—may be the real driver, so one could predict which cross-domain method will win from dataset statistics alone before running experiments.
  • The nine-submodule decomposition implies a nearly combinatorial space of cross-domain recombinations; systematic search over submodule pairings may yield additional gains beyond the PLM-ICD+Label2Vec example.
  • If the trait-over-domain conclusion generalizes, shared cross-domain benchmarks could be more informative than isolated domain-specific benchmarks, and emerging areas such as skill or occupation tagging may benefit immediately from extreme multi-label methods developed for law and e-commerce.
  • A testable extension is to check whether the surprising RoBERTa-PM-on-patents gain comes from vocabulary size rather than medical knowledge, which the paper itself leaves open.
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

6 major / 5 minor

Summary. The paper presents a cross-domain analysis of hierarchical text classification (HTC), proposing a nine-submodule unified framework to compare 32 surveyed methods, then running a unified benchmark of eight selected methods across eight datasets from five domains (legal, scientific, news, medical, patent). The central empirical claim is that state-of-the-art HTC methods often originate outside the target domain: under the unified pipeline the authors report new state-of-the-art results on NYT-166, SciHTC-83, USPTO2M-632, and MIMIC3-3681 by applying methods from other domains or combining submodules. The paper also reports dataset-characteristic correlations, a PLM-ICD+Label2Vec combination, domain-specific LLM experiments, long-document handling ablations, data-size ablations, and PLT initialization experiments. The authors release a codebase, cleaned datasets (EurLex-3985, EurLex-DC-410), and adapted versions of existing datasets.

Significance. If the empirical findings are robust, this is the first large-scale cross-domain benchmark for HTC and would be a useful contribution: the unified framework and codebase enable standardized comparison across legal, medical, patent, news, and scientific domains, and the claim that effective methods often come from other domains is practically valuable. The paper's strengths include releasing reproducible code and datasets, discussing limitations explicitly (resource caps, hyperparameter constraints, potential differences from original papers), and validating two implementations against published EurLex-3956 numbers. However, the headline 'new state-of-the-art' claims are only established relative to the authors' own reimplementations, several key margins are tiny, and the correlation analysis rests on only eight datasets without significance testing, so the more sweeping conclusions are not yet fully supported.

major comments (6)
  1. [Section 4.4, Tables 2-4, Footnote 1] The 'new state-of-the-art results' claim on NYT-166, SciHTC-83, USPTO2M-632, and MIMIC3-3681 is only supported against the eight methods reimplemented in this paper's unified pipeline. The only external validation reported is on EurLex-3956 (page 15: MatchXML 87.89/17.84 vs 88.12/-, XR-Transformer 87.83/17.82 vs 87.22/-), which is not among the four datasets where a new SOTA is claimed. Footnote 1 concedes results may differ from original papers due to preprocessing, evaluation, and implementation differences. The headline claim should either be verified against original published numbers on each SOTA-claimed dataset, or reframed as 'best under our unified reimplementation pipeline'.
  2. [Section 4.1.3, Table 1] MIMIC3-3681 and USPTO2M-632 received synthetic hierarchy levels (+3 and +2 respectively) to make them compatible with taxonomy-using methods such as THMM, HILL, and HiAGM-TP, while evaluation is on the original flat labels. The paper states this adaptation is needed but does not justify that adding artificial levels preserves task semantics. If the synthetic parent/grandparent codes are noisy or semantically inconsistent with ICD or IPC codes, methods that consume the taxonomy may be unfairly helped or harmed. An ablation (or at least an analysis of the synthetic levels' consistency) is needed before relying on rankings for these datasets.
  3. [Section 4.4, Tables 2-3] The 36-hour/40GB resource caps produce many ET/EM entries, most importantly HR-SciBERT-mt, which only ran on SciHTC-83 (and there required over a week). The claim that HR-SciBERT-mt lost SOTA on SciHTC-83 and the general claim that cross-domain methods outperform domain-native methods are therefore partially artifacts of these caps: on every other dataset the native patent/scientific/medical SOTA methods either timed out or exhausted memory. The authors should demonstrate that the main rankings are stable when the caps are relaxed for at least the domain-native SOTA methods, or explicitly restrict the cross-domain conclusions to the resource-constrained setting.
  4. [Section 4.4, Tables 2-5] All results are reported as averages over five random seeds with no standard deviations, confidence intervals, or significance tests. Several headline margins are very small: SciHTC-83 P@1 is 61.70 for MatchXML versus 61.64 for THMM; USPTO2M-632 P@1 is 82.28 for MatchXML versus 81.83 for THMM; NYT-166 P@1 is 96.64 for XR-Transformer versus 96.27 for MatchXML. Without per-seed variance, rank changes and the 'surprising' cross-domain wins may reflect noise. Please provide per-seed results or standard deviations for at least the main P@1/R@1 tables and the Table 5 combination experiments.
  5. [Section 5.1, Figure 3] The claim that 'dataset characteristics matter more than domain specificity' is largely based on Pearson correlations computed over only eight datasets (Figure 3). With n=8, an |r| of 0.535 (the reported correlation between document length and max P@1) is not statistically significant at the 0.05 level, and none of the correlations are reported with p-values or confidence intervals. The conclusions in Section 5.1 should be supported by significance testing, or presented as suggestive observations rather than confirmed findings.
  6. [Section 5.3, Figure 5] The claim that RoBERTa-PM (medical) on USPTO2M-632 surpasses SciBERT (scientific) and yields 'a surprising new state of the art' is based on a single encoder swap with no significance testing and no control for vocabulary size; the paper itself acknowledges that the improvement may stem from RoBERTa-PM's larger vocabulary rather than domain-specific knowledge. This is presented as a headline result but is not adequately separated from the alternative explanation. A vocabulary-matched control or a softened conclusion is needed.
minor comments (5)
  1. [Table 7 vs Table 2] Table 7 reports FlatBERT on USPTO2M-632 (2M) as 82.00/55.36, while Table 2 reports 81.36/54.92 for the same method and dataset; these should be reconciled or explained.
  2. [Figure 2 caption] The caption contains a typo: 'label sizses' should be 'label sizes'.
  3. [Table A1] The data augmentation method is spelled 'REMEDIAL' in Section 3.3 but 'REMIDIAL' in Table A1; the spelling should be consistent.
  4. [Section 5.6, Tables B10-B11] The PLT initialization differences (PIF A vs Gold vs Random) are all within about 0.5 percentage points and are reported without significance testing; the conclusion that 'prior knowledge of label hierarchies is somewhat beneficial' would benefit from a variance estimate.
  5. [Section 4.1.2, Table 1] The table's '(512)' notation for MIMIC3-3681 appears without explanation; it is only clarified in Appendix B.10. Please define it in the main text.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the cross-domain SOTA claims are empirical measurements under an explicitly scoped unified pipeline, not derivations that reduce to their own inputs.

full rationale

The paper is an empirical benchmark and survey, not a formal derivation, so the circularity burden is low. Its central claims—that top HTC performance can come from methods developed in other domains and that dataset characteristics matter more than domain origin—are supported by measurements under a unified evaluation pipeline, not by construction from a fitted parameter or a self-referential definition. The 'new state-of-the-art' results are explicitly scoped in footnote 1 ('These state-of-the-art results are based on our unified evaluation framework... some results may differ from those reported in original papers due to differences in preprocessing, evaluation settings, or implementation details') and in Section 4.4, and the authors validate implementation fidelity against published EurLex-3956 numbers (MatchXML 87.89/17.84 vs 88.12/-, XR-Transformer 87.83/17.82 vs 87.22/-). The only self-citations (Li, Kang, and De Bie 2023a, 2023b) appear in the concluding discussion of future work on skill tagging and are not load-bearing for any central claim. Fairness concerns raised by the skeptical review—modified hierarchies for MIMIC3-3681 and USPTO2M-632, the 36-hour/40GB resource caps, and HR-SciBERT-mt's timeouts on non-native datasets—are threats to external validity and correctness, not circularity, because they do not make the reported rankings equal to the experimental inputs by construction. No fitted parameter is renamed as a prediction, no uniqueness theorem is imported from the authors' prior work, and no result is defined in terms of the conclusion it supports. The paper's own limitations section candidly acknowledges that the unified framework 'may yield results that differ from those reported in original papers,' which further confirms that the SOTA claims are contingent empirical findings rather than self-referential derivations.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

This paper is an empirical benchmark rather than a theoretical derivation, so this ledger records the experimental design choices and dataset assumptions on which the central claims rest. No new theoretical entities are postulated.

free parameters (4)
  • Method hyperparameters per dataset = Tuned on validation splits with limited ranges
    All eight methods have hyperparameters tuned per dataset using a validation set sampled from training data (Section 4.4); the central SOTA rankings depend on these choices, and limited search may disadvantage some methods.
  • PLM-ICD segment length = Pre-defined length not specified
    PLM-ICD segments long documents into parts of pre-defined lengths (Section B); the long-document handling conclusion in Section 5.4 depends on this choice.
  • Label2Vec embedding dimension = Not reported
    The PLM-ICD+Label2Vec hybrid in Section 5.2 introduces a label-embedding module; its dimension is a design choice that affects the reported gains.
  • PLT tree initialization = PIFA, Gold, or Random
    XR-Transformer's tree initialization is a hand-chosen design variable analyzed in Section 5.6; differences are small (0.1-0.5 points) but present.
assumptions (5)
  • domain assumption The selected eight datasets represent the five domains and support general cross-domain conclusions
    Section 4.1.1 selects datasets with gold taxonomies and fewer than 4,000 labels; extreme multi-label and non-textual settings are excluded, so the conclusions may not generalize beyond this scope.
  • domain assumption Precision/Recall@k under a unified pipeline is a valid basis for comparing SOTA
    Section 4.3 defines P/R@k; many original papers report micro-F1 or other metrics, and the paper assumes the unified metric is fair across methods (Section 4.4 caveat).
  • ad hoc to paper Adding synthetic hierarchy levels to MIMIC3-3681 and USPTO2M-632 preserves task semantics
    Section 4.1.3 adds parent and grandparent labels to flat datasets to satisfy hierarchy-based methods; evaluation only uses original flat labels, but the modified hierarchy seen by methods may change their behavior.
  • domain assumption The EUROVOC taxonomy mapping for EurLex labels is correct
    Section 4.1.3 maps EurLex labels to the EUROVOC taxonomy via string similarity and manual verification; mapping errors would affect EurLex-3985 results.
  • domain assumption Correlations reported in Figure 3 are meaningful with 8 data points
    The paper reports correlations between dataset features and performance without significance tests; with N=8, these correlations are fragile and may be driven by outliers.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Your Next State-of-the-Art Could Come from Another Domain: A Cross-Domain Analysis of Hierarchical Text Classification." pith.science (2026). https://pith.science/paper/5E2QLUQ6

@misc{pith2026241212744,
  author       = {Pith},
  title        = {Pith review of: Your Next State-of-the-Art Could Come from Another Domain: A Cross-Domain Analysis of Hierarchical Text Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5E2QLUQ6}},
  note         = {Machine review of arXiv:2412.12744}
}
read the original abstract

Text classification with hierarchical labels is a prevalent and challenging task in natural language processing. Examples include assigning ICD codes to patient records, tagging patents into IPC classes, assigning EUROVOC descriptors to European legal texts, and more. Despite its widespread applications, a comprehensive understanding of state-of-the-art methods across different domains has been lacking. In this paper, we provide the first comprehensive cross-domain overview with empirical analysis of state-of-the-art methods. We propose a unified framework that positions each method within a common structure to facilitate research. Our empirical analysis yields key insights and guidelines, confirming the necessity of learning across different research areas to design effective methods. Notably, under our unified evaluation pipeline, we achieved new state-of-the-art results by applying techniques beyond their original domains.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 43 canonical work pages

  1. [1]

    , Akkaya, C

    banerjee2019hierarchical APACrefauthors Banerjee, S. , Akkaya, C. , Perez-Sorrosal, F. Tsioutsiouliklis, K. APACrefauthors \ 2019 . Hierarchical transfer learning for multi-label text classification Hierarchical transfer learning for multi-label text classification . Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics P...

    work page 2019

  2. [2]

    beltagy2019scibert APACrefauthors Beltagy, I. , Lo, K. Cohan, A. APACrefauthors \ 2019 . SciBERT: A pretrained language model for scientific text Scibert: A pretrained language model for scientific text . arXiv preprint arXiv:1903.10676 ,

    arXiv 2019

  3. [3]

    u l, F. , Hauss, F. , Menkel, L. , Sch \

    bertalis2024hierarchical APACrefauthors Bertalis, N. , Granse, P. , G \"u l, F. , Hauss, F. , Menkel, L. , Sch \"u ler, D. Scherp, A. APACrefauthors \ 2024 . Hierarchical Text Classification (HTC) vs. eXtreme Multilabel Classification (XML): Two Sides of the Same Medal Hierarchical text classification (htc) vs. extreme multilabel classification (xml): Two...

    work page arXiv 2024

  4. [4]

    , Dahiya, K

    Bhatia16extreme APACrefauthors Bhatia, K. , Dahiya, K. , Jain, H. , Kar, P. , Mittal, A. , Prabhu, Y. Varma, M. APACrefauthors \ 2016 . The extreme classification repository: Multi-label datasets and code. The extreme classification repository: Multi-label datasets and code. APACrefURL http://manikvarma.org/downloads/XC/XMLRepository.html APACrefURL

    work page 2016

  5. [5]

    , Todorovski, L

    bogatinovski2022comprehensive APACrefauthors Bogatinovski, J. , Todorovski, L. , D z eroski, S. Kocev, D. APACrefauthors \ 2022 . Comprehensive comparative study of multi-label classification methods Comprehensive comparative study of multi-label classification methods . Expert Systems with Applications 203 117215,

    work page 2022

  6. [6]

    , Fergadiotis, E

    chalkidis2019extreme APACrefauthors Chalkidis, I. , Fergadiotis, E. , Malakasiotis, P. , Aletras, N. Androutsopoulos, I. APACrefauthors \ 2019 . Extreme Multi-Label Legal Text Classification: A Case Study in EU Legislation Extreme multi-label legal text classification: A case study in eu legislation . Proceedings of the Natural Legal Language Processing W...

    work page 2019

  7. [7]

    , Yu, H F

    chang2020taming APACrefauthors Chang, W C. , Yu, H F. , Zhong, K. , Yang, Y. Dhillon, I.S. APACrefauthors \ 2020 . Taming pretrained transformers for extreme multi-label text classification Taming pretrained transformers for extreme multi-label text classification . Proceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data ...

    work page 2020

  8. [8]

    , Rivera, A

    charte2014concurrence APACrefauthors Charte, F. , Rivera, A. , del Jesus, M.J. Herrera, F. APACrefauthors \ 2014 . Concurrence among imbalanced labels and its influence on multilabel resampling algorithms Concurrence among imbalanced labels and its influence on multilabel resampling algorithms . Hybrid Artificial Intelligence Systems: 9th International Co...

    work page 2014

Show all 53 references
  1. [9]

    , Rivera, A.J

    charte2019dealing APACrefauthors Charte, F. , Rivera, A.J. , del Jesus, M.J. Herrera, F. APACrefauthors \ 2019 . Dealing with difficult minority labels in imbalanced mutilabel data sets Dealing with difficult minority labels in imbalanced mutilabel data sets . Neurocomputing 3...

    2019

  2. [10]

    , Huang, X

    chen2020hyperbolic APACrefauthors Chen, B. , Huang, X. , Xiao, L. , Cai, Z. Jing, L. APACrefauthors \ 2020 . Hyperbolic interaction model for hierarchical multi-label classification Hyperbolic interaction model for hierarchical multi-label classification . Proceedings of the A...

    2020

  3. [11]

    , Cheng, J

    chen2022survey APACrefauthors Chen, X. , Cheng, J. , Liu, J. , Xu, W. , Hua, S. , Tang, Z. Sheng, V.S. APACrefauthors \ 2022 . A survey of multi-label text classification based on deep learning A survey of multi-label text classification based on deep learning . International ...

    2022

  4. [12]

    APACrefauthors \ 2018

    devlin2018bert APACrefauthors Devlin, J. APACrefauthors \ 2018 . Bert: Pre-training of deep bidirectional transformers for language understanding Bert: Pre-training of deep bidirectional transformers for language understanding . arXiv preprint arXiv:1810.04805 ,

    2018 arXiv

  5. [13]

    , Junge, A

    edin2023automated APACrefauthors Edin, J. , Junge, A. , Havtorn, J.D. , Borgholt, L. , Maistro, M. , Ruotsalo, T. Maal e, L. APACrefauthors \ 2023 . Automated medical coding on MIMIC-III and MIMIC-IV: a critical review and replicability study Automated medical coding on mimic-...

    2023

  6. [14]

    , Wadbude, R

    gupta2019distributional APACrefauthors Gupta, V. , Wadbude, R. , Natarajan, N. , Karnick, H. , Jain, P. Rai, P. APACrefauthors \ 2019 . Distributional semantics meets multi-label learning Distributional semantics meets multi-label learning . Proceedings of the AAAI Conference ...

    2019

  7. [15]

    han2023survey APACrefauthors Han, M. , Wu, H. , Chen, Z. , Li, M. Zhang, X. APACrefauthors \ 2023 . A survey of multi-label classification based on supervised and semi-supervised learning A survey of multi-label classification based on supervised and semi-supervised learning ....

    2023

  8. [16]

    , Tsai, S C

    huang2022plm APACrefauthors Huang, C W. , Tsai, S C. Chen, Y N. APACrefauthors \ 2022 . PLM-ICD: Automatic ICD coding with pretrained language models Plm-icd: Automatic icd coding with pretrained language models . arXiv preprint arXiv:2207.05289 ,

    2022 arXiv

  9. [17]

    , Chen, E

    huang2019hierarchical APACrefauthors Huang, W. , Chen, E. , Liu, Q. , Chen, Y. , Huang, Z. , Liu, Y. Wang, S. APACrefauthors \ 2019 . Hierarchical multi-label text classification: An attention-based recurrent network approach Hierarchical multi-label text classification: An at...

    2019

  10. [18]

    , Wang, D

    jiang2022exploiting APACrefauthors Jiang, T. , Wang, D. , Sun, L. , Chen, Z. , Zhuang, F. Yang, Q. APACrefauthors \ 2022 . Exploiting Global and Local Hierarchies for Hierarchical Text Classification Exploiting global and local hierarchies for hierarchical text classification ...

    2022

  11. [19]

    , Pollard, T.J

    johnson2016mimic APACrefauthors Johnson, A.E. , Pollard, T.J. , Shen, L. , Lehman, L w.H. , Feng, M. , Ghassemi, M. Mark, R.G. APACrefauthors \ 2016 . MIMIC-III, a freely accessible critical care database Mimic-iii, a freely accessible critical care database . Scientific data ...

    2016

  12. [20]

    , Salampasis, M

    kamateri2024will APACrefauthors Kamateri, E. , Salampasis, M. Perez-Molina, E. APACrefauthors \ 2024 . Will AI solve the patent classification problem? Will ai solve the patent classification problem? World Patent Information 78 102294,

    2024

  13. [21]

    , Xiao, H

    khandagale2020bonsai APACrefauthors Khandagale, S. , Xiao, H. Babbar, R. APACrefauthors \ 2020 . Bonsai: diverse and shallow trees for extreme multi-label classification Bonsai: diverse and shallow trees for extreme multi-label classification . Machine Learning 109 11 2099--2119,

    2020

  14. [22]

    , Gupta, D

    kharbanda2024gandalf APACrefauthors Kharbanda, S. , Gupta, D. , Schultheis, E. , Banerjee, A. , Hsieh, C J. Babbar, R. APACrefauthors \ 2024 . Gandalf: Learning Label-label Correlations in Extreme Multi-label Classification via Label Features Gandalf: Learning label-label corr...

    2024

  15. [23]

    , Brown, D.E

    kowsari2017hdltex APACrefauthors Kowsari, K. , Brown, D.E. , Heidarysafa, M. , Meimandi, K.J. , Gerber, M.S. Barnes, L.E. APACrefauthors \ 2017 . Hdltex: Hierarchical deep learning for text classification Hdltex: Hierarchical deep learning for text classification . 2017 16th I...

    2017

  16. [24]

    \ Hsiang, J

    lee2020patent APACrefauthors Lee, J S. \ Hsiang, J. APACrefauthors \ 2020 . Patent classification by fine-tuning BERT language model Patent classification by fine-tuning bert language model . World Patent Information 61 101965,

    2020

  17. [25]

    , Ott, M

    lewis2020pretrained APACrefauthors Lewis, P. , Ott, M. , Du, J. Stoyanov, V. APACrefauthors \ 2020 . Pretrained language models for biomedical and clinical tasks: understanding and extending the state-of-the-art Pretrained language models for biomedical and clinical tasks: und...

    2020

  18. [26]

    , Yadav, V

    li2024scalable APACrefauthors Li, D. , Yadav, V. , Zhu, Z.L. , Fard, M.M. , Afzal, Z. Tsatsaronis, G. APACrefauthors \ 2024 . Scalable Patent Classification with Aggregated Multi-View Ranking Scalable patent classification with aggregated multi-view ranking . Proceedings of th...

    2024

  19. [27]

    , Kang, B

    li2023llm4jobs APACrefauthors Li, N. , Kang, B. De Bie, T. APACrefauthors \ 2023 1 . LLM4Jobs: Unsupervised occupation extraction and standardization leveraging Large Language Models Llm4jobs: Unsupervised occupation extraction and standardization leveraging large language mod...

    2023 arXiv

  20. [28]

    , Kang, B

    li2023skillgpt APACrefauthors Li, N. , Kang, B. De Bie, T. APACrefauthors \ 2023 2 . SkillGPT: a RESTful API service for skill extraction and standardization using a Large Language Model Skillgpt: a restful api service for skill extraction and standardization using a large lan...

    2023 arXiv

  21. [29]

    li2018deeppatent APACrefauthors Li, S. , Hu, J. , Cui, Y. Hu, J. APACrefauthors \ 2018 . DeepPatent: patent classification with convolutional neural networks and word embedding Deeppatent: patent classification with convolutional neural networks and word embedding . Scientomet...

    2018

  22. [30]

    , Liang, W

    liu2023recent APACrefauthors Liu, R. , Liang, W. , Luo, W. , Song, Y. , Zhang, H. , Xu, R. Liu, M. APACrefauthors \ 2023 . Recent Advances in Hierarchical Multi-label Text Classification: A Survey Recent advances in hierarchical multi-label text classification: A survey . arXi...

    2023 arXiv

  23. [31]

    \ F \"u rnkranz, J

    loza2008efficient APACrefauthors Loza Menc \' a, E. \ F \"u rnkranz, J. APACrefauthors \ 2008 . Efficient pairwise multilabel classification for large-scale problems in the legal domain Efficient pairwise multilabel classification for large-scale problems in the legal domain ....

    2008

  24. [32]

    , Tian, J

    mao2019hierarchical APACrefauthors Mao, Y. , Tian, J. , Han, J. Ren, X. APACrefauthors \ 2019 . Hierarchical Text Classification with Reinforced Label Assignment Hierarchical text classification with reinforced label assignment . Proceedings of the 2019 Conference on Empirical...

    2019

  25. [33]

    , Sachdeva, N

    mittal2021eclare APACrefauthors Mittal, A. , Sachdeva, N. , Agrawal, S. , Agarwal, S. , Kar, P. Varma, M. APACrefauthors \ 2021 . ECLARE: Extreme classification with label graph correlations Eclare: Extreme classification with label graph correlations . Proceedings of the Web ...

    2021

  26. [34]

    , Gross, S

    paszke2019pytorch APACrefauthors Paszke, A. , Gross, S. , Massa, F. , Lerer, A. , Bradbury, J. , Chanan, G. others APACrefauthors \ 2019 . Pytorch: An imperative style, high-performance deep learning library Pytorch: An imperative style, high-performance deep learning library ...

    2019

  27. [35]

    , Friedrich, A

    pujari2021multi APACrefauthors Pujari, S.C. , Friedrich, A. Str \"o tgen, J. APACrefauthors \ 2021 . A multi-task approach to neural multi-label hierarchical patent classification using transformers A multi-task approach to neural multi-label hierarchical patent classification...

    2021

  28. [36]

    \ Caragea, C

    sadat2022hierarchical APACrefauthors Sadat, M. \ Caragea, C. APACrefauthors \ 2022 . Hierarchical Multi-Label Classification of Scientific Documents Hierarchical multi-label classification of scientific documents . Proceedings of the 2022 Conference on Empirical Methods in Nat...

    2022

  29. [37]

    , Jain, A.K

    saini2021galaxc APACrefauthors Saini, D. , Jain, A.K. , Dave, K. , Jiao, J. , Singh, A. , Zhang, R. Varma, M. APACrefauthors \ 2021 . GalaXC: Graph neural networks with labelwise attention for extreme classification Galaxc: Graph neural networks with labelwise attention for ex...

    2021

  30. [38]

    APACrefauthors \ 2008

    sandhaus2008new APACrefauthors Sandhaus, E. APACrefauthors \ 2008 . The new york times annotated corpus The new york times annotated corpus . Linguistic Data Consortium, Philadelphia 6 12 e26752,

    2008

  31. [39]

    , Giacobini, M

    tarekegn2021review APACrefauthors Tarekegn, A.N. , Giacobini, M. Michalak, K. APACrefauthors \ 2021 . A review of methods for imbalanced multi-label classification A review of methods for imbalanced multi-label classification . Pattern Recognition 118 107965,

    2021

  32. [40]

    , Ullah, M

    tarekegn2024deep APACrefauthors Tarekegn, A.N. , Ullah, M. Cheikh, F.A. APACrefauthors \ 2024 . Deep learning for multi-label learning: A comprehensive survey Deep learning for multi-label learning: A comprehensive survey . arXiv preprint arXiv:2401.16549 ,

    2024 arXiv

  33. [41]

    wang2021cognitive APACrefauthors Wang, B. , Hu, X. , Li, P. Philip, S.Y. APACrefauthors \ 2021 . Cognitive structure learning model for hierarchical multi-label text classification Cognitive structure learning model for hierarchical multi-label text classification . Knowledge-...

    2021

  34. [42]

    , Wang, P

    wang2022incorporating APACrefauthors Wang, Z. , Wang, P. , Huang, L. , Sun, X. Wang, H. APACrefauthors \ 2022 . Incorporating Hierarchy into Text Encoder: a Contrastive Learning Approach for Hierarchical Text Classification Incorporating hierarchy into text encoder: a contrast...

    2022

  35. [43]

    , Mao, Z

    wei2022survey APACrefauthors Wei, T. , Mao, Z. , Shi, J X. , Li, Y F. Zhang, M L. APACrefauthors \ 2022 . A survey on extreme multi-label learning A survey on extreme multi-label learning . arXiv preprint arXiv:2210.03968 ,

    2022 arXiv

  36. [44]

    , Teng, S

    xu2021hierarchical APACrefauthors Xu, L. , Teng, S. , Zhao, R. , Guo, J. , Xiao, C. , Jiang, D. Ren, B. APACrefauthors \ 2021 . Hierarchical multi-label text classification with horizontal and vertical category correlations Hierarchical multi-label text classification with hor...

    2021

  37. [45]

    , Jha, K

    xun2020correlation APACrefauthors Xun, G. , Jha, K. , Sun, J. Zhang, A. APACrefauthors \ 2020 . Correlation networks for extreme multi-label text classification Correlation networks for extreme multi-label text classification . Proceedings of the 26th ACM SIGKDD International ...

    2020

  38. [46]

    , Sunderraman, R

    ye2024matchxml APACrefauthors Ye, H. , Sunderraman, R. Ji, S. APACrefauthors \ 2024 . MatchXML: An Efficient Text-label Matching Framework for Extreme Multi-label Text Classification Matchxml: An efficient text-label matching framework for extreme multi-label text classificati...

    2024

  39. [47]

    , Zhang, Z

    you2019attentionxml APACrefauthors You, R. , Zhang, Z. , Wang, Z. , Dai, S. , Mamitsuka, H. Zhu, S. APACrefauthors \ 2019 . Attentionxml: Label tree-based attention-aware deep model for high-performance extreme multi-label text classification Attentionxml: Label tree-based att...

    2019

  40. [48]

    , Chang, W C

    zhang2021fast APACrefauthors Zhang, J. , Chang, W C. , Yu, H F. Dhillon, I. APACrefauthors \ 2021 . Fast multi-resolution transformer fine-tuning for extreme multi-label text classification Fast multi-resolution transformer fine-tuning for extreme multi-label text classificati...

    2021

  41. [49]

    zhang2022hcn APACrefauthors Zhang, X. , Xu, J. , Soh, C. Chen, L. APACrefauthors \ 2022 . La-hcn: label-based attention for hierarchical multi-label text classification neural network La-hcn: label-based attention for hierarchical multi-label text classification neural network...

    2022

  42. [50]

    zhou2020hierarchy APACrefauthors Zhou, J. , Ma, C. , Long, D. , Xu, G. , Ding, N. , Zhang, H. Liu, G. APACrefauthors \ 2020 . Hierarchy-aware global model for hierarchical text classification Hierarchy-aware global model for hierarchical text classification . Proceedings of th...

    2020

  43. [51]

    zhu2024hill APACrefauthors Zhu, H. , Wu, J. , Liu, R. , Hou, Y. , Yuan, Z. , Li, S. Xu, K. APACrefauthors \ 2024 . HILL: Hierarchy-aware Information Lossless Contrastive Learning for Hierarchical Text Classification Hill: Hierarchy-aware information lossless contrastive learni...

    2024

  44. [52]

    write newline

    " write newline " cite write " FUNCTION editor.postfix editor num.names #1 > "( )" "( )" if FUNCTION editor.trans.postfix editor num.names #1 > "( )" "( )" if FUNCTION trans.postfix translator num.names #1 > "( )" "( )" if FUNCTION authors.editors.reflist.apa5 'field := 'dot :...

  45. [53]

    write newline

    " write newline "" before.all 'output.state := FUNCTION string.to.integer 't := t text.length 'k := #1 'char.num := t char.num #1 substring 's := s is.num s "." = or char.num k = not and char.num #1 + 'char.num := while char.num #1 - 'char.num := t #1 char.num substring FUNCTI...

Pith tools

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