Pith. sign in

REVIEW 3 major objections 6 minor 27 references

Classifying and Tracking International Aid Contribution Towards SDGs

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The paper claims a multi-label classifier trained with SDG, country, and LLM signals reaches F1 0.843 on aid projects and reveals hidden 2015–2022 financing trends when applied to CRS records.

desk verdict A useful classifier wrapped around a trend analysis that does not identify what it claims; the classification part deserves a serious referee, the trend part needs rework. read the letter →

arxiv 2505.15223 v2 pith:HBKSY3A6 submitted 2025-05-21 cs.CY

Sungwon Park , Dongjoon Lee , Kyeongjin Ahn , Yubin Choi , Junho Lee , Meeyoung Cha , Kyung Ryul Park This is my paper · ORCID
classification cs.CY
keywords multi-labeltextclassificationSustainableDevelopmentGoalsCreditorReportingSystemofficialassistanceLLMknowledgeinjectioncontrastivelearningaidfinancingtrendsexpertevaluation
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 tries to make the Creditor Reporting System (CRS) tractable for SDG tracking: it trains a multi-label classifier that assigns any subset of the 17 Sustainable Development Goals to an aid project from its text description, donor, and recipient. The authors claim the model outperforms both LLM-based and graph-based baselines, reaching F1 0.843 and AUROC 0.9617, and that each of its three modules — SDG-semantics injection, country-pair attention, and gated LLM decisions — adds measurable value. They then apply the model to 1.7 million CRS records from 2015–2022, imputing the SDG labels that were never recorded, and use a regression fit to estimate how annual aid budgets split across goals. If the classification holds, the payoff is practical: agencies can automate SDG tagging, reduce subjective reporting, and see financing trends that were previously hidden by missing data. The paper also reports expert interviews endorsing the tool's efficiency and standardization value.

What carries the argument

The load-bearing object is the calibrated fused representation $h^{cls}_x = \hat{h}_x + \alpha_x \bar{h}_x$, where $\hat{h}_x$ is the country-attentive representation formed by cross-attention over donor–recipient policy summaries and $\bar{h}_x$ is the LLM-decision representation formed by cross-attention over goal embeddings for the LLM's predicted label set; $\alpha_x$ is a learned usefulness score that predicts whether the LLM decision overlaps the true label set. Complementing this fusion are the Semantics Injection module, which builds positive contrastive samples by Gumbel-softmax token sampling aligned to the official SDG definitions, and two auxiliary binary cross-entropy losses on the country and LLM attentive representations. Together these components guide the encoder to focus on goal-relevant text while weighing external knowledge sources.

What would settle it

Re-run the Eq. 21 budget fit with 2023 data added and with nonnegative or L2-regularized weights; if the estimated per-SDG budget proportions move by more than a few percentage points, the reported trends—such as the COVID spike in SDG 3—are artifacts of the underdetermined fitting procedure.

Watch

Extended reading notes

Core claim

The central claim is that a multi-label aid-classification model, built on a multilingual BERT encoder and guided by three auxiliary signals, can label CRS projects with SDG goals more accurately than prior methods, and that the resulting labels, applied to the full 2015–2022 CRS corpus, reveal longitudinal financing shifts that are invisible in the labeled subset alone. The model injects official SDG definitions into the encoder via contrastive learning with goal-conditioned token sampling, adds a cross-attention module that conditions on donor and recipient country policy summaries, and appends an LLM decision representation gated by a learned usefulness score. On the test set the full model reports F1 0.8430 and AUROC 0.9617, with each module contributing in ablations; the LLM gate especially improves recall for underrepresented goals such as SDG 7 and SDG 14. Applying the model to 1,719,733 records from 2015 to 2022 lets the authors estimate per-SDG budget proportions, yielding trends like a near-doubling of SDG 3 funding during the COVID-19 pandemic and distinct allocation patterns across recipient income groups.

Load-bearing premise

The analysis of aid financing presumes that 17 per-SDG budget weights can be recovered from the five yearly budget totals in Eq. 21 without regularization or constraints—an underdetermined fit in which the estimated budget trends hinge on arbitrary solver choices.

Editorial extensions

If this is right

  • If the reported F1 of 0.843 transfers to production, aid agencies can cut manual SDG tagging time while keeping or increasing review accuracy.
  • The imputed 2015–2022 labels make retroactive financing analysis possible: budget shares per SDG, across income groups and donors, can be estimated for years when the CRS recorded no SDG data.
  • The estimated budget trends, such as the near-doubling of SDG 3 funding during the COVID-19 peak and the income-gradient shift from SDG 1 and 2 toward SDG 8 and 9 in middle-income countries, would provide a data foundation for policy planning.
  • The modular design suggests a template for multi-label classification with label semantics, country context, and LLM prior knowledge that could generalize beyond aid records.
  • If the approach is adopted, it could strengthen the case for making the CRS SDG field mandatory, since automated classification lowers the reporting burden on donor agencies.

Reading between the lines

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

  • Editorial extension: the financing trends should be treated as hypotheses until the Eq. 21 fit is shown to be stable under regularization or with additional years; the paper's own setup (five observations, seventeen weights) leaves the budget estimates underdetermined.
  • Editorial extension: the same three-module recipe could be applied to other under-specified multi-label tasks with official definitions, such as tagging scientific papers with SDGs or classifying corporate sustainability reports.
  • Editorial extension: the usefulness gate $\alpha_x$ could double as a model-confidence signal, flagging projects where the classifier and LLM disagree for human review, which is a natural next step for deployment.
  • Editorial extension: releasing the imputed 1.7-million-record dataset as a benchmark would let the research community audit per-SDG recall and measure how much the label imputation changes when new labeled data arrive.
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

3 major / 6 minor

Summary. The paper presents a multi-label text classification framework that assigns Sustainable Development Goal (SDG) labels to international aid projects in the OECD CRS database. The method combines a BERT-style encoder with three auxiliary modules: SDG semantics injection via contrastive learning, country-information-guided attention, and an LLM-decision-guided module with a learned usefulness weight. The authors report that their model outperforms several baselines on a held-out test set (F1 0.8430, AUROC 0.9617) and then apply the trained model to impute missing SDG labels for roughly one million CRS records from 2015-2022. Using these imputed labels, they fit a linear regression (Eq. 21) to estimate per-SDG budget weights and claim to reveal temporal trends such as a COVID-19-related spike in SDG 3 funding. The paper also reports expert interviews suggesting practical utility. The central contributions are the classifier itself and the longitudinal financing analysis that claims to uncover hidden trends in aid allocation.

Significance. If the classification results are robust, the model would be a practically useful tool for aid agencies and for completing the voluntary SDG focus index in CRS; the authors have released code and partnered with government agencies, which strengthens the applied relevance. The expert interviews provide qualitative evidence of usability. However, the longitudinal financing analysis is the paper's headline 'hidden trends' contribution, and it is currently not identified by the data. The classification evaluation also lacks statistical rigor, with a single split and small margins over baselines. The paper is therefore a promising application with a load-bearing methodological weakness in the trend analysis.

major comments (3)
  1. [Sec. 5.1, Eq. (21)] The optimization problem in Eq. (21) estimates 17 weights w_k from only five yearly observations (i ∈ {2018,...,2022}), so the 5×17 design matrix has rank at most 5 and the least-squares objective is underdetermined. With no stated regularization, constraints, or prior, any solver returns an arbitrary solution (e.g., a minimum-norm solution); consequently the budget proportions in Table 6 and the trend shapes in Figures 3, 4, and 6 are not identified by the data. The paper's claim that the model 'can reveal hidden trends' is therefore unsupported; the COVID-era SDG 3 spike and other patterns could disappear under an equally valid solution. The limitation paragraph in Sec. 5.1 acknowledges missing labels but does not acknowledge this identification failure.
  2. [Sec. 4.1, Table 4] The performance comparison rests on a single 3:1 split with no confidence intervals, standard deviations, or significance tests. The margins over the best graph-based baselines are small (F1 0.8430 vs 0.8365 for LGCN and 0.8357 for GCLR; AUROC 0.9617 vs 0.9521 and 0.9519). Without repeated splits or statistical testing, the statement that the framework 'consistently surpasses baseline models across all metrics' is not established, and the observed differences may be within run-to-run variation.
  3. [Sec. 5.1 and Figure 4] The longitudinal analysis applies the fitted (and unidentified) weights to statistics c_i^k that are computed from model-imputed labels for all projects, including 382,083 pre-2018 records and 616,726 post-2018 records with missing labels. Because per-SDG performance is markedly lower for underrepresented goals (Figure 7, e.g., SDGs 7, 12, 14, 15), the imputed inputs to Eq. (21) carry substantial classifier error; the reported trends therefore conflate financing changes with classification artifacts. The in-paper limitation concedes that incomplete labels 'may not accurately reflect the actual aid allocations,' but the subsequent conclusions in Sec. 5.1 do not hedge accordingly.
minor comments (6)
  1. [Eq. (5) and Eq. (3) text] The text mentions 'thresold' and 'Gumble-Softmax'; these should be 'threshold' and 'Gumbel-Softmax.'
  2. [Table 4] The entry 'Vallia BCE' should be 'Vanilla BCE.'
  3. [Sec. 3.2 and Sec. 3.3] There are typos: 'doner-recipient' should be 'donor-recipient' and 'databse' should be 'database.'
  4. [Eq. (12)] The product notation \prod_{j\in\bar{y}} h_{g_j} is ambiguous because \bar{y} is a set of goal indices and the product of embedding vectors is not defined; please specify the intended combination (e.g., element-wise product) and its ordering.
  5. [Sec. 4.1 Implementation] The phrase 'trained LLMs over 100 epochs' is ambiguous; clarify whether the LLM backbone is frozen or fine-tuned, and reconcile this with Table 7's sensitivity analysis of different LLM backbones.
  6. [Figure 4] Figure 4 plots estimates for 2016-2017 even though Eq. (21) is fit only to 2018-2022; the procedure used to produce the earlier points should be stated explicitly.

Circularity Check

1 steps flagged · score 6.0 of 10

The classifier evaluation is self-contained, but the longitudinal budget trends in Sec. 5.1 reduce to the fitted weights of an underdetermined regression (Eq. 21) applied to model-imputed labels.

  1. fitted input called prediction [Section 5.1 (Analysis of Aid Financing Using the Trained Model), Eq. 21; Table 6; Figures 3, 4, and 6]
    "We assume that the annual budget sum S_i for the i-th year in the labeled dataset can be expressed as a linear sum of the statistics of the i-th year for each k-th goal, c_i^k, within the labeled dataset. To estimate the budget w_k for each k-th goal, we fit a linear regression model by solving the optimization problem (Eq. 21). The estimated budget proportions, detailed in the Appendix Table 6, are then used in the financing analysis."

    The 'estimated budget proportions' in Table 6 are exactly the fitted parameters w_k of Eq. 21. Since Eq. 21 has only five yearly observations (2018-2022) but 17 unknown weights, the least-squares problem is underdetermined and the reported proportions are an arbitrary solver-dependent solution rather than quantities identified by the data. The financing analysis then multiplies these same fitted weights by label statistics that, for 2015-2017 and for missing records, are predictions of the same model, producing Figures 3, 4, and 6. The highlighted 'hidden trends,' including the SDG 3 COVID spike, are therefore in-sample transformations of the fitted weights and model-imputed inputs, not independent measurements; the fitted parameters are presented as an empirical finding.

full rationale

The core multi-label classification contribution is not circular: the model is trained and evaluated on real SDG focus labels from the CRS, with a held-out test set, and the reported F1 and AUROC improvements over baselines are independent empirical results. The self-citations to the authors' prior work (Lee et al. 2023; Park and Park 2024) are contextual and not load-bearing for the main claims. The only step that reduces to its own inputs is the longitudinal financing analysis in Sec. 5.1: Eq. 21 fits 17 per-SDG budget weights to just five annual budget sums, and the resulting budget proportions and trend figures are the fitted weights applied to model-imputed label statistics. This makes the paper's secondary claim about 'hidden trends' partially circular, or more precisely an unidentifiable re-display of fitted quantities. The classification results themselves remain externally grounded, so the overall circularity score is moderate rather than total.

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

The classifier rests on standard supervised-learning assumptions plus the use of voluntary SDG labels as ground truth. The budget trend analysis additionally requires an underdetermined linear fit (Eq. 21) and the model's own imputed labels, which is the main source of circularity and instability.

free parameters (3)
  • budget weight per SDG w_k = e.g., SDG 8: 1.00, SDG 3: 0.97, SDG 4: 0.02 (Table 6)
    Fitted via Eq. 21 from 5 annual observations to 17 unknowns; underdetermined and used to produce all budget trend estimates.
  • loss weights lambda_1, lambda_2 = 0.1
    Set by hand in Section 4.1; no sensitivity analysis reported in main text.
  • token sampling threshold tau = 0.01
    Chosen by hand in Section 4.1 for positive sample construction in Eq. 5.
assumptions (3)
  • domain assumption Voluntary SDG focus index labels are accurate enough to serve as ground truth for training.
    Section 1 and 4.1 note the labels are voluntary, incomplete, and subjective; the classifier is trained on them as if they were correct.
  • domain assumption Project descriptions and donor/recipient codes contain enough signal to determine SDG alignment.
    The model classifies using only these features; if descriptions are too broad, the task may be ill-posed.
  • ad hoc to paper Annual budget can be decomposed as a linear sum of per-goal statistics with fixed weights (Eq. 21).
    Introduced specifically for the financing analysis; no justification or validation is given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Classifying and Tracking International Aid Contribution Towards SDGs." pith.science (2026). https://pith.science/paper/HBKSY3A6

@misc{pith2026250515223,
  author       = {Pith},
  title        = {Pith review of: Classifying and Tracking International Aid Contribution Towards SDGs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HBKSY3A6}},
  note         = {Machine review of arXiv:2505.15223}
}
read the original abstract

International aid is a critical mechanism for promoting economic growth and well-being in developing nations, supporting progress toward the Sustainable Development Goals (SDGs). However, tracking aid contributions remains challenging due to labor-intensive data management, incomplete records, and the heterogeneous nature of aid data. Recognizing the urgency of this challenge, we partnered with government agencies to develop an AI model that complements manual classification and mitigates human bias in subjective interpretation. By integrating SDG-specific semantics and leveraging prior knowledge from language models, our approach enhances classification accuracy and accommodates the diversity of aid projects. When applied to a comprehensive dataset spanning multiple years, our model can reveal hidden trends in the temporal evolution of international development cooperation. Expert interviews further suggest how these insights can empower policymakers with data-driven decision-making tools, ultimately improving aid effectiveness and supporting progress toward SDGs.

Figures

Figures reproduced from arXiv: 2505.15223 by the authors.

Figure 1
Figure 1. The proposed framework has three main components: (1) [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Distribution of SDG focus index by frequency ( [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Distribution of Estimated SDG Budget Allocation by GNI per Capita. SDGs with a ratio of less than 1% have been omitted. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Estimated annual budget allocation. Over a 7-year period [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Example Summarization of LLM Development Policy [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Estimated Annual Budget Allocation for Each SDG. The full dataset, which includes past and missing data, exhibits a different [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Detailed performance metrics, including F1-score, precision, and recall, for each SDG. [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 26 canonical work pages

  1. [1]

    Linkages between official development assistance and the sustainable devel- opment goals: a scoping review.International Policy Cen- tre for Inclusive Growth, page 446,

    [Alsayyad, 2020] Amina Said Alsayyad. Linkages between official development assistance and the sustainable devel- opment goals: a scoping review.International Policy Cen- tre for Inclusive Growth, page 446,

    work page 2020

  2. [3]

    Aid allocation and poverty reduction.European Economic Review, 46(8):1475–1500,

    [Collier and Dollar, 2002] Paul Collier and David Dollar. Aid allocation and poverty reduction.European Economic Review, 46(8):1475–1500,

    work page 2002

  3. [5]

    These categorizations, which are fun- damental to the CRS database and international development policy frameworks, revealed the language model’s ability to construct coherent, contextually appropriate profiles based on recipient income classifications. For donor nations, the model showed a reasonable under- standing of basic policy orientations and commo...

    work page 2023

  4. [13]

    Aid quality and donor rankings.World Development, 39(11):1907–1917,

    [Knacket al., 2011 ] Stephen Knack, F Halsey Rogers, and Nicholas Eubank. Aid quality and donor rankings.World Development, 39(11):1907–1917,

    work page 2011

  5. [15]

    Machine learning driven aid classification for sustainable development

    [Leeet al., 2023 ] Junho Lee, Hyeonho Song, Dongjoon Lee, Sundong Kim, JiSoo Sim, Meeyoung Cha, and Kyung- Ryul Park. Machine learning driven aid classification for sustainable development. InProc. of the IJCAI, pages 6040–6048,

    work page 2023

  6. [17]

    Foreign assistance: An introduction to US programs and policy.Congressional Research Service, 10,

    [Morgenstern and Brown, 2022] Emily M Morgenstern and Nick M Brown. Foreign assistance: An introduction to US programs and policy.Congressional Research Service, 10,

    work page 2022

  7. [19]

    Addressing institutional challenges in sustainable development goals implementation: Lessons from the re- public of korea.Sustainable Development, 32(1):1354– 1369,

    [Park and Park, 2024] Kyung Ryul Park and Young Shil Park. Addressing institutional challenges in sustainable development goals implementation: Lessons from the re- public of korea.Sustainable Development, 32(1):1354– 1369,

    work page 2024

  8. [21]

    Linking aid to the sustainable development goals–A machine learning approach

    [Pincetet al., 2019 ] Arnaud Pincet, Shu Okabe, and Martin Pawelczyk. Linking aid to the sustainable development goals–A machine learning approach

    work page 2019

Show all 27 references
  1. [22]

    Classifier chains for multi-label classification

    [Readet al., 2009 ] Jesse Read, Bernhard Pfahringer, Geoff Holmes, and Eibe Frank. Classifier chains for multi-label classification. InIn proc. of the ECML-PKDD, pages 254–

    2009

  2. [24]

    Incorporating hi- erarchy into text encoder: a contrastive learning approach for hierarchical text classification

    [Wanget al., 2022 ] Zihan Wang, Peiyi Wang, Lianzhe Huang, Xin Sun, and Houfeng Wang. Incorporating hi- erarchy into text encoder: a contrastive learning approach for hierarchical text classification. InProc. of the ACL, pages 7109–7119,

    2022

  3. [25]

    Label-specific document representation for multi-label text classification

    [Xiaoet al., 2019 ] Lin Xiao, Xin Huang, Boli Chen, and Liping Jing. Label-specific document representation for multi-label text classification. InProc. of the EMNLP- IJCNLP, pages 466–475,

    2019

  4. [26]

    Link prediction based on graph neural networks.Advances in Neurips, 31,

    [Zhang and Chen, 2018] Muhan Zhang and Yixin Chen. Link prediction based on graph neural networks.Advances in Neurips, 31,

    2018

  5. [2002]

    On the empirics of foreign aid and growth.Economic Journal, 114:191–191, 02

    [Dalgaardet al., 2004 ] Carl-Johan Dalgaard, Henrik Hansen, and Finn Tarp. On the empirics of foreign aid and growth.Economic Journal, 114:191–191, 02

    2004

  6. [2004]

    BERT: Pre-training of deep bidirectional transformers for language understand- ing.arXiv preprint arXiv:1810.04805,

    [Devlinet al., 2018 ] Jacob Devlin, Ming-Wei Chang, Ken- ton Lee, and Kristina Toutanova. BERT: Pre-training of deep bidirectional transformers for language understand- ing.arXiv preprint arXiv:1810.04805,

    2018 arXiv

  7. [2008]

    Are aid agencies improv- ing?Economic Policy, 22(52):634–678, 08

    [Easterly, 2014] William Easterly. Are aid agencies improv- ing?Economic Policy, 22(52):634–678, 08

    2014

  8. [2009]

    OECD publishing,

    [UNAIDS, 2004] UNAIDS.Creditor Reporting System on Aid Activities Aid Activities in Support of HIV/AIDS Con- trol Volume 2004 Issue 6: Activit ´es d’aide pour la lutte contre le VIH/SIDA Volume 2004-6. OECD publishing,

    2004

  9. [2011]

    Classification of goods using text descriptions with sen- tences retrieval.arXiv preprint arXiv:2111.01663,

    [Leeet al., 2021 ] Eunji Lee, Sundong Kim, Sihyun Kim, Sungwon Park, Meeyoung Cha, Soyeon Jung, Suyoung Yang, Yeonsoo Choi, Sungdae Ji, Minsoo Song, et al. Classification of goods using text descriptions with sen- tences retrieval.arXiv preprint arXiv:2111.01663,

    2021 arXiv

  10. [2013]

    Categorical reparametrization with gumble-softmax

    [Janget al., 2017 ] Eric Jang, Shixiang Gu, and Ben Poole. Categorical reparametrization with gumble-softmax. In Proc. of ICLR,

    2017

  11. [2014]

    [Ericsson and Mealy, 2019] Fredrik Ericsson and Sam Mealy. Connecting official development assistance and science technology and innovation for inclusive devel- opment: Measurement challenges from a Development Assistance Committee perspective

    2019

  12. [2017]

    A results agenda for the 2030 agenda: New approaches for changing contexts

    [Kasneci, 2018] Ola Kasneci. A results agenda for the 2030 agenda: New approaches for changing contexts

    2018

  13. [2018]

    Where does the money go? best and worst prac- tices in foreign aid.Journal of Economic Perspectives, 22(2):29–52, June

    [Easterly and Pfutze, 2008] William Easterly and Tobias Pfutze. Where does the money go? best and worst prac- tices in foreign aid.Journal of Economic Perspectives, 22(2):29–52, June

    2008

  14. [2019]

    Simcse: Simple contrastive learning of sentence embeddings

    [Gaoet al., 2021 ] T Gao, X Yao, and Danqi Chen. Simcse: Simple contrastive learning of sentence embeddings. In Proc. of the EMNLP,

    2021

  15. [2020]

    A simple framework for contrastive learning of visual representations

    [Chenet al., 2020 ] Ting Chen, Simon Kornblith, Moham- mad Norouzi, and Geoffrey Hinton. A simple framework for contrastive learning of visual representations. InProc. of the ICML, pages 1597–1607,

    2020

  16. [2021]

    The evolution of official development assistance: Achieve- ments, criticisms and a way forward

    [Hynes and Scott, 2013] William Hynes and Simon Scott. The evolution of official development assistance: Achieve- ments, criticisms and a way forward

    2013

  17. [2022]

    Proposal to include an SDG focus field in the CRS database

    [OECD, 2018] OECD. Proposal to include an SDG focus field in the CRS database. DCD/DAC/STAT(2018)1,

    2018

  18. [2023]

    Label-specific dual graph neural network for multi-label text classification

    [Maet al., 2021 ] Qianwen Ma, Chunyuan Yuan, Wei Zhou, and Songlin Hu. Label-specific dual graph neural network for multi-label text classification. InProc. of the ACL- IJCNLP, pages 3855–3864,

    2021

  19. [2024]

    Knowledge sharing via domain adaptation in customs fraud detection

    [Parket al., 2022 ] Sungwon Park, Sundong Kim, and Meey- oung Cha. Knowledge sharing via domain adaptation in customs fraud detection. InProc. of the AAAI, volume 36, pages 12062–12070,

    2022

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.