Pith. sign in

REVIEW 5 major objections 6 minor 43 references

This paper proposes SDG polarity detection—judging whether a news text signals progress, regression, or neutrality toward a specific Sustainable Development Goal—and introduces SDG-POD, a benchmark for it.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-04 15:19 UTC pith:K7VPKUQ6

load-bearing objection SDG-POD is a real contribution, but the synthetic-data claim is confounded and the test-set noise makes the headline gaps unsupported as reported. the 5 major comments →

arxiv 2509.19833 v4 pith:K7VPKUQ6 submitted 2025-09-24 cs.CL cs.AIcs.DL

Polarity Detection of Sustainable Development Goals in News Text

Andrea Cadeddu , Alessandro Chessa , Vincenzo De Leo , Gianni Fenu , Francesco Osborne , Diego Reforgiato Recupero , Angelo Salatino , Luca Secchi This is my paper
classification cs.CL cs.AIcs.DL
keywords SDG polarity detectionSDG-POD benchmarklarge language modelssynthetic data augmentationmulti-LLM majority votingzero-shot vs fine-tuningsustainability monitoringtext classification
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper introduces SDG-POD, a benchmark for a new task: given a news text and one of the UN's 17 Sustainable Development Goals, decide whether the text reports progress toward that goal, regression away from it, or neither. The central claim is that the task is genuinely hard for current open-source large language models, but that fine-tuning on a training set labelled by majority vote of five LLMs consistently improves performance. The best model, QWQ-32B, reaches a macro F1 of 61.6, up from 57.8 in zero-shot, and importantly reduces the severe error of swapping positive and negative labels. The paper also shows that synthetic training data can substitute for scarce human annotations in this domain. If the benchmark holds up, it gives sustainability researchers a reusable testbed and a low-cost recipe for building polarity detectors.

Core claim

The paper's central discovery is that SDG polarity—whether a text reports progress toward, regression from, or neutrality about a given goal—can be learned from synthetic training data produced by majority-voting independent LLM annotations, improving on zero-shot prompting. Fine-tuned QWQ-32B reaches macro F1 61.6 versus 57.8 zero-shot; the best zero-shot model, Phi-4, scores 59.8. The gain is concentrated where it matters: fine-tuning reduces positive/negative label confusion, and under error-weighted F1 the gap widens from 1–2 points to 10–11 points. Per-SDG gains are largest on SDG-9, SDG-12, and SDG-15. The task remains challenging, making SDG-POD a demanding benchmark.

What carries the argument

SDG-POD, a benchmark of 6,400 texts: 5,824 automatically labelled by a majority vote of five different LLMs and 576 independently labelled by human annotators. The load-bearing mechanism is the multi-LLM majority-voting pipeline that converts noisy individual predictions into silver training labels, combined with a cost-sensitive F1 evaluation that penalises positive/negative confusion.

Load-bearing premise

The 576 human-labelled test items are treated as ground truth; if those labels are unreliable—and the paper reports Cohen's Kappa as low as 0.38 among its annotators—then every model ranking and F1 gap inherits that unreliability.

What would settle it

Take the SDG-POD test set, have a fresh panel of annotators label the same 576 texts under the same guidelines, and re-run the comparison. If pairwise agreement on the shared items stays below 0.6 and the fine-tuned QWQ-32B no longer beats zero-shot Phi-4, the paper's central claim of consistent fine-tuning gains is not stable.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Fine-tuning on synthetic multi-LLM data beats zero-shot prompting: QWQ-32B moves from 57.8 to 61.6 macro F1, and Phi-4 from 59.8 to 61.3.
  • Fine-tuning reduces the most consequential errors: positive/negative label swaps are largely replaced by neutral misclassifications, and the error-weighted F1 gap widens to 10–11 points at high penalties.
  • SDG polarity is separable from sentiment: a celebratory tone can describe regression, and a bitter tone can describe progress, so sentiment classifiers cannot be reused directly.
  • The SDG-POD benchmark is reusable and open, providing a testbed for future models on a task where human annotations are scarce.
  • Zero-shot performance varies only modestly across models (57.7–59.8 F1), so the task is not yet saturated and synthetic fine-tuning is a practical route for deployment.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Editorial extension: because the test set is English and drawn from formal reports and policy documents, the polarity signal may not transfer to informal or multilingual text; a multilingual extension would reveal whether the task's difficulty is linguistic or conceptual.
  • Editorial extension: the reported 1.8-point gap between fine-tuned QWQ-32B and zero-shot Phi-4 sits close to the inter-annotator agreement range (Kappa 0.38–0.69), so a confidence-interval or noise-aware analysis is the natural next step for deciding whether the gap is stable.
  • Editorial extension: the same multi-LLM vote-and-fine-tune recipe could be applied to other low-resource policy-monitoring tasks, such as detecting whether corporate climate pledges represent progress or greenwashing.
  • Editorial extension: because synthetic-data fine-tuning reduces positive/negative confusion, a testable next step is to compare against a same-size human-labelled training set; similar or better F1 would isolate the value of annotation quality versus scale.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The paper introduces a new task, SDG polarity detection (positive/neutral/negative with respect to a stated SDG), and presents SDG-POD, a 6,400-text benchmark built from OSDG excerpts. Training labels are produced by a five-LLM ensemble with majority voting and hand-written tie-break rules; the 576-item test set is annotated by six human annotators in two groups, with only 48 shared texts per group for agreement estimation. Six open-source models are evaluated in zero-shot and fine-tuned settings, and the paper claims that fine-tuning on the synthetic training data consistently improves performance, that QWQ-32B is the best model (macro F1 61.6), and that fine-tuning reduces critical positive/negative confusions, especially for SDG-9, SDG-12, and SDG-15.

Significance. If the central claims hold, the paper would provide a reusable evaluation harness and a low-cost LLM-voting annotation recipe for a resource-constrained, socially important domain. The authors are to be credited for releasing the codebase, for making the task definition explicit, and for honestly reporting the low inter-annotator agreement. However, the main empirical claims -- that synthetic training data improves robustness and that QWQ-32B is the best fine-tuned model -- are currently not established by the evidence as presented, because the experimental design confounds fine-tuning with label source, the comparison is not controlled, and the human-annotated test set is too noisy and too thinly annotated for the claimed differences to be reliable.

major comments (5)
  1. [§4.3, Tables 1-2] The test-set reliability is the load-bearing uncertainty. Cohen's Kappa values range from 0.38 to 0.60 in Group 1 and 0.55 to 0.69 in Group 2, yet these estimates come from only 48 shared texts per group. Moreover, 480 of the 576 test items are annotated by a single evaluator, with no reliability estimate for that majority. Every reported metric -- the 1.8-point gap between the best fine-tuned and best zero-shot model, the per-SDG rankings, and the confusion-matrix claims in Tables 5-6 -- inherits this label noise. The paper needs bootstrap confidence intervals, noise-aware evaluation, or a reliability analysis that treats the gold labels as uncertain; otherwise the stated conclusions are unsupported.
  2. [§5, §6] The claim that synthetic data improves performance is confounded. The paper's hypothesis is that if fine-tuned models outperform their non-fine-tuned counterparts, this would indicate that the synthetic training data is valuable. But this comparison differs in two variables at once: fine-tuning itself and the provenance of the training labels. There is no control fine-tuned on non-synthetic (e.g., human-annotated) labels, nor a control using single-LLM labels, so the observed gains cannot be attributed to the synthetic label-generation method. Similarly, the claim that fine-tuning reduces critical positive/negative errors is based on comparing Phi-4 zero-shot (Table 5) with QWQ-32B fine-tuned (Table 6), which confounds model identity with training setting.
  3. [§6, Tables 4 and 7] Phi-4 was fine-tuned for 10 epochs whereas all other models were fine-tuned for 5 epochs, explicitly 'in order to improve its classification metrics.' This makes cross-model comparisons of fine-tuned F1 scores unreliable and directly affects the reported 1.8-point gap between Phi-4 and QWQ-32B. If the goal is to compare model families under a standard protocol, the same number of epochs must be used or the results must be shown to be robust to epoch choice. The current presentation violates the standardised fine-tuning procedure stated at the start of §5.
  4. [§6, Figure 1 and Tables 5-6] The per-SDG analysis is based on only 36 test examples per SDG, and the authors draw strong conclusions about SDG-specific rankings (e.g., SDG-9, SDG-12, SDG-15) without any uncertainty quantification. With a 3-class setup and 36 examples, a difference of a few instances can move a per-SDG F1 score by several points. Bootstrap intervals or error bars are necessary before per-SDG strengths can be claimed. The confusion matrices also need clearer formatting and row/column sums; as printed, several cell values do not obviously sum to the stated class totals.
  5. [§4.2, §6] The training labels themselves are generated by the authors' own five-LLM ensemble with hand-written Bronze tie-break rules, and there is no validation of these synthetic labels against human judgments. The paper presents no evidence that the ensemble's labels are accurate enough to serve as a training signal, beyond the aggregate downstream comparison that is already confounded. At minimum, a sample-based human evaluation of the synthetic labels, or a comparison with training on human labels of the same texts, is needed to support the claim that the proposed annotation recipe produces a 'high-quality resource for fine-tuning.'
minor comments (6)
  1. [Title and Abstract] The title of the full text reads 'Polarity Detection of Sustainable Detection Goals,' which should be 'Sustainable Development Goals.' The abstract and title also say 'News Text,' but Section 4.1 describes the source as OSDG excerpts from reports, policy documents, and publication abstracts; please clarify the scope.
  2. [Figure 1 caption] The caption refers to 'Phi4-8B,' but the model evaluated in the paper is PHI4-4B. Please correct the inconsistency.
  3. [§6, paragraph after Table 4] The sentence 'with a measured F1-score of 61.6%, an increase of 3.8 percentage points compared to the performance of the non-fine-tuned version of the same model' is ambiguous and appears to attribute the 61.6% figure to Phi-4 when it is actually QWQ-32B's score. Please rewrite to state clearly which model each number refers to.
  4. [§2] There is a typo: 'named olarity detection' should be 'named polarity detection.' Also, the related-work section would benefit from a clearer distinction between the proposed SDG-polarity task and the existing 'polarization analysis' literature, since the terms are used almost interchangeably.
  5. [§4.3] The description of the test-set annotation says a majority vote was used for the 48 shared texts, but it is not stated how the final label was assigned in the 2 'Bronze' cases (three different labels). The text mentions that neutral was chosen, but this appears only after the distribution of shared labels and should be stated more prominently.
  6. [Tables 5-6] The confusion matrices are difficult to parse because the column headers are repeated across three blocks and the cell values are not visually aligned. Please reformat as standard confusion matrices and verify all row sums, as some cells appear inconsistent with the stated class sizes.

Circularity Check

0 steps flagged

No significant circularity: the synthetic-data claim is empirically evaluated against an independently human-annotated test set.

full rationale

The paper's central claim is that fine-tuning on LLM-generated synthetic training data improves SDG polarity detection. The training labels are produced by a five-LLM majority vote (Section 4.2), but the test set is independently annotated by six human evaluators (Section 4.3). The improvement is measured by comparing zero-shot and fine-tuned versions of the same six models on that human test set (Tables 3–7). No equation or construction forces the result: the synthetic labels are not fitted to the test labels, and the test labels are not derived from the training pipeline. The paper does not rely on self-citations to justify its central premise; references to prior work are standard model/dataset citations. Low inter-annotator agreement (kappa 0.38–0.69) is a validity concern about label noise, but it is not a circularity: the evaluation would still be an external test even with noisy labels. Consequently, no circular step meeting the quoted-evidence standard is present.

Axiom & Free-Parameter Ledger

4 free parameters · 4 axioms · 1 invented entities

The central claims rest on four unverified premises: OSDG's SDG association is correct; human polarity labels are reliable (contradicted in Group 1 by the paper's own Kappa of 0.38); LLM-voted labels carry valid training signal; and the single-label three-way formulation is coherent. Free choices include the Bronze tie-break rules, the cost weights used to amplify the robustness finding, Phi-4's extra five epochs, and the 36-text per-SDG test cells. The only invented object is the SDG polarity construct itself, and it has a measurable, though noisy, handle in the human test set. No physical entities, mediators, or dimensions are introduced.

free parameters (4)
  • Bronze tie-break rules for training labels = 2pos+2neg+1neu -> neutral; 2pos+2neu+1neg -> positive; 2neg+2neu+1pos -> negative
    Ad hoc decision rules for the 283 training texts with no majority among five annotator LLMs (Section 4.2). The rule that a positive/negative tie resolves to neutral shapes the training distribution by hand and biases contested items toward the majority class.
  • Error-weight costs in weighted F1 = cost_pn = cost_np = 1, 2, 10
    The robustness claim ('fine-tuning reduces critical errors') relies on hand-chosen penalties applied post hoc to confusion matrices (Section 6, Table 7); at cost=1 the fine-tuning gain is only 1-2 F1 points.
  • Phi-4 training epochs = 10 (all other models: 5)
    Section 6 states Phi-4 was fine-tuned for 10 epochs 'in order to improve its classification metrics', i.e., the training schedule was adjusted after inspecting test-set performance.
  • Per-SDG test cell size = 36 texts per SDG
    The test set is split into 36 texts per SDG (Section 4.3), and per-SDG conclusions (SDG-9/12/15) are drawn from cells this small with no confidence intervals.
axioms (4)
  • domain assumption OSDG dataset SDG associations for the 6,400 sampled excerpts are correct
    SDG-POD inherits each text's SDG label from the OSDG community dataset (Section 4.1); if the original association is wrong, the polarity label is defined relative to the wrong goal. No re-validation of SDG association is reported.
  • domain assumption Human polarity annotations constitute reliable ground truth
    All metrics are computed against the human-labeled test set (Section 4.3). The paper's own reliability statistics contradict strong reliability: pairwise Cohen's Kappa as low as 0.38 in Group 1.
  • domain assumption LLM majority-vote labels are valid training signal
    All 5,824 training labels come from five LLMs with heuristic tie-breaking (Section 4.2); no validation of training-label accuracy against human labels is reported, and the generator models could be systematically wrong in the same direction.
  • domain assumption Each text has a single unambiguous polarity with respect to an SDG
    The task is formalized as single-label, three-class classification (Section 3). The Kappa range of 0.38-0.69 suggests the single-label assumption is contested by the annotators' own disagreements, and the paper's own Bronze cases had to be resolved by fiat.
invented entities (1)
  • SDG polarity construct (positive/neutral/negative with respect to a specific SDG) independent evidence
    purpose: Defines the new classification target that all models are measured against and that the benchmark operationalizes
    The construct is operationalized through annotation guidelines and the 576-text human test set, giving a measurable handle outside the paper. The handle is weak, however: the authors' own Cohen's Kappa is as low as 0.38, so the independent evidence is noisy.

pith-pipeline@v1.3.0-alltime-deepseek · 17747 in / 19907 out tokens · 127846 ms · 2026-08-04T15:19:40.394983+00:00 · methodology

0 comments
read the original abstract

The United Nations' Sustainable Development Goals (SDGs) provide a globally recognised framework for addressing major societal, environmental, and economic challenges. While recent advances in natural language processing (NLP) and large language models (LLMs) have enabled the automatic identification of SDG-related content, they do not capture whether the described events represent progress toward or regression from a specific goal. To address this gap, we introduce the novel task of SDG polarity detection and present SDG-POD, a benchmark dataset combining manually annotated and synthetically generated examples. We evaluate six state-of-the-art open-source LLMs under both zero-shot and fine-tuning settings and investigate the impact of synthetic data augmentation on model performance. Our results show that SDG polarity detection remains challenging for current LLMs; however, fine-tuned models, particularly QWQ-32B, achieve the best overall performance, with especially strong results on SDG-9, SDG-12, and SDG-15. Furthermore, we demonstrate that synthetic training data consistently improves model robustness and classification performance. This work introduces a new benchmark for SDG polarity detection and provides practical insights into developing LLM-based systems for sustainability monitoring.

Figures

Figures reproduced from arXiv: 2509.19833 by Alessandro Chessa, Andrea Cadeddu, Angelo Salatino, Diego Reforgiato Recupero, Francesco Osborne, Gianni Fenu, Luca Secchi, Vincenzo De Leo.

Figure 1
Figure 1. Figure 1: F1 comparison by SDG for QWQ-32B and Phi4-8B LLMs. [PITH_FULL_IMAGE:figures/full_fig_p022_1.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

43 extracted references · 13 canonical work pages

  1. [1]

    Hirai, F

    T. Hirai, F. Comim, Measuring the sustainable development goals: A poset analysis, Ecological Indicators 145 (2022) 109605.doi:https: //doi.org/10.1016/j.ecolind.2022.109605. URLhttps://www.sciencedirect.com/science/article/pii/S147 0160X22010780

    arXiv 2022

  2. [2]

    Ziegler, S

    D. Ziegler, S. Wolff, A.-B. Agu, G. Cortiana, M. Umair, F. de Durfort, E. Neumann, G. Walther, J. Kristiansen, M. Lienkamp, How to Measure Sustainability? An Open-Data Approach, Sustainability 15 (4) (2023) 3203–. URLhttps://ideas.repec.org/a/gam/jsusta/v15y2023i4p3203-d 1063481.html

    2023

  3. [3]

    Mishra, S

    M. Mishra, S. Desul, C. A. G. Santos, S. K. Mishra, A. H. M. Ka- mal, S. Goswami, A. M. Kalumba, R. Biswal, R. M. da Silva, C. A. C. dos Santos, K. Baral, A bibliometric analysis of sustainable develop- ment goals (sdgs): a review of progress, challenges, and opportuni- 27 ties, Environment, Development and Sustainability (2023) 1–43doi: 10.1007/s10668-02...

    work page doi:10.1007/s10668-023-03225-w 2023

  4. [4]

    La Fleur, Using large language models to help train machine learning sdg classifiers, Working Paper 180, UN DESA, New York (November 2023)

    M. La Fleur, Using large language models to help train machine learning sdg classifiers, Working Paper 180, UN DESA, New York (November 2023)

    2023

  5. [5]

    Nilashi, O

    M. Nilashi, O. Keng Boon, G. Tan, B. Lin, R. Abumalloh, Critical Data Challenges in Measuring the Performance of Sustainable Development Goals: Solutions and the Role of Big-Data Analytics, Harvard Data Science Review 5 (3), https://hdsr.mitpress.mit.edu/pub/9n4uzkg3 (jul 27 2023)

    2023

  6. [6]

    J. E. Guisiano, R. Chiky, J. De Mello, Sdg-meter: A deep learning based tool for automatic text classification of the sustainable development goals, in: N. T. Nguyen, T. K. Tran, U. Tukayev, T.-P. Hong, B. Traw- iński, E. Szczerbicki (Eds.), Intelligent Information and Database Sys- tems, Springer International Publishing, Cham, 2022, pp. 259–271

    2022

  7. [7]

    Matsui, K

    T. Matsui, K. Suzuki, K. Ando, Y. Kitai, C. Haga, N. Masuhara, S. Kawakubo, A natural language processing model for supporting sus- tainable development goals: translating semantics, visualizing nexus, and connecting stakeholders, Sustainability Science 17 (3) (2022) 969– 985.doi:10.1007/s11625-022-01093-3. URLhttps://doi.org/10.1007/s11625-022-01093-3

    work page doi:10.1007/s11625-022-01093-3 2022

  8. [8]

    Guisiano, R

    J. Guisiano, R. Chiky, Automatic classification of multilabel texts re- latedtosustainabledevelopmentgoals(sdgs), in: TECHENVEGC2021, Montpellier, France, 2021. URLhttps://hal.archives-ouvertes.fr/hal-03154261

    2021

  9. [9]

    Bianchi, M

    F. Bianchi, M. Marelli, P. Nicoli, M. Palmonari, SWEAT: Scoring po- larization of topics across different corpora, in: M.-F. Moens, X. Huang, L. Specia, S. W.-t. Yih (Eds.), Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, Association for Computational Linguistics, Online and Punta Cana, Dominican Repub- lic, 2021, p...

    work page doi:10.18653/v1/2021.emnlp-main.788 2021

  10. [10]

    R.Németh, Ascopingreviewontheuseofnaturallanguageprocessingin researchonpoliticalpolarization: trendsandresearchprospects, Journal of Computational Social Science 6 (1) (2023) 289–313.doi:10.1007/ s42001-022-00196-2

    2023

  11. [11]

    B. Pang, L. Lee, S. Vaithyanathan, Thumbs up? sentiment classification using machine learning techniques, in: Proceedings of the ACL-02 Con- ference on Empirical Methods in Natural Language Processing - Volume 10, EMNLP ’02, Association for Computational Linguistics, USA, 2002, p. 79–86.doi:10.3115/1118693.1118704. URLhttps://doi.org/10.3115/1118693.1118704

    arXiv 2002

  12. [12]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, I. Polosukhin, Attention is all you need, in: Proceedings of the 31st International Conference on Neural Information Processing Systems, NIPS’17, Curran Associates Inc., Red Hook, NY, USA, 2017, p. 6000–6010

    2017

  13. [14]

    Brown, et al., Language models are few-shot learners, in: H

    T. Brown, et al., Language models are few-shot learners, in: H. Larochelle, M. Ranzato, R. Hadsell, M. Balcan, H. Lin (Eds.), Ad- vances in Neural Information Processing Systems, Vol. 33, Curran As- sociates, Inc., 2020, pp. 1877–1901. URLhttps://proceedings.neurips.cc/paper_files/paper/2020/ file/1457c0d6bfcb4967418bfb8ac142f64a-Paper.pdf

    2020

  14. [15]

    K. L. Tan, C. P. Lee, K. M. Lim, A survey of sentiment analysis: Ap- proaches, datasets, and future research, Applied Sciences 13 (7) (2023) 4550.doi:10.3390/app13074550

    work page doi:10.3390/app13074550 2023

  15. [16]

    P. Xia, S. Wu, B. V. Durme, Which *bert? a survey organizing con- textualized encoders., in: B. Webber, T. Cohn, Y. He, Y. Liu (Eds.), 29 EMNLP (1), Association for Computational Linguistics, 2020, pp. 7516– 7533. URLhttp://dblp.uni-trier.de/db/conf/emnlp/emnlp2020-1.htm l#XiaWD20

    2020

  16. [17]

    Alaparthi, M

    S. Alaparthi, M. Mishra, Bert: a sentiment analysis odyssey, Journal of Marketing Analytics 9 (2) (2021) 118–126, published on June 1, 2021. doi:10.1057/s41270-021-00109-8. URLhttps://doi.org/10.1057/s41270-021-00109-8

    work page doi:10.1057/s41270-021-00109-8 2021

  17. [18]

    Batra, N

    H. Batra, N. S. Punn, S. K. Sonbhadra, S. Agarwal, Bert-based sen- timent analysis: A software engineering perspective, in: C. Strauss, G. Kotsis, A. M. Tjoa, I. Khalil (Eds.), Database and Expert Systems Applications, Springer International Publishing, Cham, 2021, pp. 138– 148

    2021

  18. [19]

    Y. Wu, Z. Jin, C. Shi, P. Liang, T. Zhan, Research on the application of deep learning-based bert model in sentiment analysis, Applied and Computational Engineering 67 (2024) 280–286.doi:10.54254/2755-2 721/67/2024MA

    work page doi:10.54254/2755-2 2024

  19. [20]

    S. T. Kokab, S. Asghar, S. Naz, Transformer-based deep learning models for the sentiment analysis of social media data, Array 14 (2022) 100157. doi:10.1016/j.array.2022.100157

    arXiv 2022

  20. [21]

    Achiam, et al., Gpt-4 technical report (2024).arXiv:2303 .08774

    OpenAI, J. Achiam, et al., Gpt-4 technical report (2024).arXiv:2303 .08774. URLhttps://arxiv.org/abs/2303.08774

    Pith/arXiv arXiv 2024

  21. [22]

    Grattafiori, et al., The llama 3 herd of models (2024).arXiv:2407 .21783

    A. Grattafiori, et al., The llama 3 herd of models (2024).arXiv:2407 .21783. URLhttps://arxiv.org/abs/2407.21783

    Pith/arXiv arXiv 2024

  22. [23]

    J. O. Krugmann, J. Hartmann, Sentiment analysis in the age of gener- ative ai, Customer Needs and Solutions 11 (2024) 3.doi:10.1007/s4 0547-024-00143-4

    doi:10.1007/s4 2024

  23. [24]

    Küçük, F

    D. Küçük, F. Can, Stance detection: Concepts, approaches, resources, and outstanding issues, in: Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Re- trieval, SIGIR ’21, Association for Computing Machinery, New York, 30 NY, USA, 2021, p. 2673–2676.doi:10.1145/3404835.3462815. URLhttps://doi.org/10.1145/...

    arXiv 2021

  24. [26]

    Alturayeif, H

    N. Alturayeif, H. Luqman, M. Ahmed, A systematic review of machine learning techniques for stance detection and its applications, Neural Computing and Applications 35 (7) (2023) 5113–5144, epub 2023 Jan 28.doi:10.1007/s00521-023-08285-7

    doi:10.1007/s00521-023-08285-7 2023

  25. [27]

    ALDayel, W

    A. ALDayel, W. Magdy, Stance detection on social media: State of the art and trends, Information Processing & Management 58 (4) (2021) 102597.doi:https://doi.org/10.1016/j.ipm.2021.102597. URLhttps://www.sciencedirect.com/science/article/pii/S030 6457321000960

    arXiv 2021

  26. [28]

    Chriqui, I

    A. Chriqui, I. Yahav, Hebert and hebemo: A hebrew bert model and a tool for polarity analysis and emotion recognition, INFORMS Journal on Data Science 1 (1) (2022) 81–95.doi:10.1287/ijds.2022.0016. URLhttp://dx.doi.org/10.1287/ijds.2022.0016

    arXiv 2022

  27. [29]

    S. Gode, S. Bare, B. Raj, H. Yoo, Understanding political polarization using language models: A dataset and method, AI Magazine 44 (3) (2023) 248–254.arXiv:https://onlinelibrary.wiley.com/doi/pd f/10.1002/aaai.12104,doi:https://doi.org/10.1002/aaai.121 04. URLhttps://onlinelibrary.wiley.com/doi/abs/10.1002/aaai.1 2104

    work page doi:10.1002/aaai.12104 2023

  28. [30]

    M. Y. Kabir, S. K. Madria, A deep learning approach for ideology detection and polarization analysis using covid-19 tweets, in: Lecture Notes in Computer Science, vol. 13607, Springer, 2022, pp. 209–223. doi:10.1007/978-3-031-17995-2\_15

    work page doi:10.1007/978-3-031-17995-2 2022

  29. [31]

    Badjatiya, S

    P. Badjatiya, S. Gupta, M. Gupta, V. Varma, Deep learning for hate speech detection in tweets, in: Proceedings of the 26th International Conference on World Wide Web Companion, WWW ’17 Companion, 31 International World Wide Web Conferences Steering Committee, Re- public and Canton of Geneva, CHE, 2017, p. 759–760.doi:10.1145/ 3041021.3054223. URLhttps://d...

    arXiv 2017

  30. [32]

    Schmidt-Traub, C

    G. Schmidt-Traub, C. Kroll, K. Teksoz, D. Durand-Delacre, J. D. Sachs, National baselines for the sustainable development goals assessed in the sdg index and dashboards, Nature Geoscience 10 (8) (2017) 547–555. doi:10.1038/ngeo2985. URLhttps://doi.org/10.1038/ngeo2985

    work page doi:10.1038/ngeo2985 2017

  31. [33]

    Vanderfeesten, R

    M. Vanderfeesten, R. Jaworek, L. Keßler, Ai for mapping multi-lingual academic papers to the united nations’ sustainable development goals (sdgs) (1.0), Zenodo (2022).doi:10.5281/zenodo.6487606. URLhttps://doi.org/10.5281/zenodo.6487606

    work page doi:10.5281/zenodo.6487606 2022

  32. [34]

    Pradhan, L

    P. Pradhan, L. Costa, D. Rybski, W. Lucht, J. P. Kropp, A systematic study of sustainable development goal (sdg) interactions, Earth’s Future 5 (11) (2017) 1169–1179.arXiv:https://agupubs.onlinelibrary.wi ley.com/doi/pdf/10.1002/2017EF000632,doi:https://doi.org/10 .1002/2017EF000632. URLhttps://agupubs.onlinelibrary.wiley.com/doi/abs/10.100 2/2017EF000632

    work page doi:10.1002/2017ef000632 2017

  33. [35]

    Rosenberg, C

    E. Rosenberg, C. Tarazona, F. Mallor, H. Eivazi, D. Pastor-Escuredo, F. Fuso-Nerini, R. Vinuesa, Sentiment analysis on twitter data towards climate action, Results in Engineering 19 (2023) 101287.doi:https: //doi.org/10.1016/j.rineng.2023.101287. URLhttps://www.sciencedirect.com/science/article/pii/S259 0123023004140

    arXiv 2023

  34. [36]

    Gennari, M

    P. Gennari, M. D’Orazio, A statistical approach for assessing progress towards the sdg targets, Statistical Journal of the IAOS 36 (4) (2020) 1129–1142.arXiv:https://journals.sagepub.com/doi/pdf/10.32 33/SJI-200688,doi:10.3233/SJI-200688. URLhttps://journals.sagepub.com/doi/abs/10.3233/SJI-20068 8

    work page doi:10.3233/sji-200688 2020

  35. [37]

    nlp for sdgs: Measuringcorporatealignmentwiththesustainabledevelopmentgoals

    M. Chen, G. Mussalli, A. Amel-Zadeh, M. O. Weinberg, "nlp for sdgs: Measuringcorporatealignmentwiththesustainabledevelopmentgoals", 32 The Journal of Impact and ESG Investing 2 (3), spring 2022 (2022). doi:10.3905/jesg.2021.1.035

    work page doi:10.3905/jesg.2021.1.035 2022

  36. [38]

    C. Funk, E. Tönjes, R. Teuber, L. Breuer, Finding common develop- ment paths in voluntary national reviews reporting on sustainable de- velopment goals using aspect-based sentiment analysis, PLoS One 19 (8) (2024) e0307886, published on August 14, 2024.doi:10.1371/journa l.pone.0307886

    doi:10.1371/journa 2024

  37. [39]

    A. Q. Jiang, A. Sablayrolles, A. Roux, A. Mensch, B. Savary, C. Bam- ford, D. S. Chaplot, D. de las Casas, E. B. Hanna, F. Bressand, G. Lengyel, G. Bour, G. Lample, L. R. Lavaud, L. Saulnier, M.-A. Lachaux, P. Stock, S. Subramanian, S. Yang, S. Antoniak, T. L. Scao, T. Gervet, T. Lavril, T. Wang, T. Lacroix, W. E. Sayed, Mixtral of experts (2024).arXiv:2401.04088

    Pith/arXiv arXiv 2024

  38. [40]

    Abdin, et al., Phi-3 technical report: A highly capable language model locally on your phone (2024).arXiv:2404.14219

    M. Abdin, et al., Phi-3 technical report: A highly capable language model locally on your phone (2024).arXiv:2404.14219

    Pith/arXiv arXiv 2024

  39. [41]

    G. Team, T. Mesnard, et al., Gemma: Open models based on gemini research and technology (2024).arXiv:2403.08295. URLhttps://arxiv.org/abs/2403.08295

    Pith/arXiv arXiv 2024

  40. [42]

    Qwen, :, A. Yang, B. Yang, B. Zhang, B. Hui, B. Zheng, B. Yu, C. Li, D. Liu, F. Huang, H. Wei, H. Lin, J. Yang, J. Tu, J. Zhang, J. Yang, J. Yang, J. Zhou, J. Lin, K. Dang, K. Lu, K. Bao, K. Yang, L. Yu, M. Li, M. Xue, P. Zhang, Q. Zhu, R. Men, R. Lin, T. Li, T. Tang, T. Xia, X. Ren, X. Ren, Y. Fan, Y. Su, Y. Zhang, Y. Wan, Y. Liu, Z. Cui, Z. Zhang, Z. Qi...

    Pith/arXiv arXiv 2025

  41. [43]

    Devlin, M

    J. Devlin, M. Chang, K. Lee, K. Toutanova, BERT: pre-training of deep bidirectional transformers for language understanding, in: J. Burstein, C. Doran, T. Solorio (Eds.), Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL-HLT 2019, Min- neapolis, MN, USA, Ju...

    doi:10.18653/v1/n19-1423 2019

  42. [44]

    Mehra, R

    S. Mehra, R. Louka, Y. Zhang, ESGBERT: language model to help with classification tasks related to companies environmental, social, and governance practices, CoRR abs/2203.16788 (2022).arXiv:2203.167 88,doi:10.48550/ARXIV.2203.16788. URLhttps://doi.org/10.48550/arXiv.2203.16788

    Pith/arXiv arXiv doi:10.48550/arxiv.2203.16788 2022

  43. [45]

    S. A. P. Parambath, N. Usunier, Y. Grandvalet, Optimizing f- measures by cost-sensitive classification, in: Z. Ghahramani, M. Welling, C. Cortes, N. Lawrence, K. Weinberger (Eds.), Advances in Neural In- formation Processing Systems, Vol. 27, Curran Associates, Inc., 2014. URLhttps://proceedings.neurips.cc/paper_files/paper/2014/ file/5c0314ec1b57fcd36bbb...

    2014