Finnish SQuAD: A Simple Approach to Machine Translation of Span Annotations

Emil Nuutinen; Filip Ginter; Iiro Rastas

REVIEW 4 major objections 6 minor 26 references

Finnish SQuAD: A Simple Approach to Machine Translation of Span Annotations

T0 review · 4 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a span-annotated question-answering dataset can be machine-translated by coloring the answer spans and letting a commercial translation service preserve the formatting, producing a Finnish SQuAD2.0 that keeps 97.2%…

desk verdict A practical, low-cost trick for span-preserving MT of QA datasets, honestly evaluated; the Finnish resource is useful even if the headline advantage is partly self-referential. read the letter →

arxiv 2501.05963 v1 pith:NCPYKCGY submitted 2025-01-10 cs.CL

Emil Nuutinen , Iiro Rastas , Filip Ginter This is my paper · ORCID

classification cs.CL

keywords FinnishSQuADmachinetranslationspanannotationsDeepLquestionansweringdatasetannotationtransfer

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 sets out to show that the hardest part of translating a question-answering dataset, keeping the answer spans aligned with the translated passages, can be avoided entirely by encoding the spans as colored text and asking a commercial machine-translation service to translate a formatted document. Applied to SQuAD2.0, the method preserves 97.2% of the original question-answer pairs and produces a Finnish dataset on which a standard Finnish BERT reader reaches F1 73.7, compared with 61.9 on the previous translate-and-align Finnish SQuAD. If the claim holds, any span-annotated dataset can be ported to any language supported by the service for roughly the cost of the API calls, without hand-written alignment heuristics.

What carries the argument

The mechanism is DeepL's formatting-preserving document translation, treated as a span-transfer channel. Input documents encode each answer as a distinct color, with overlapping answers split so the shared region gets its own color; after translation the colors are mapped back to answer spans, trailing punctuation is stripped, and discontinuous spans produced by translation are stored both as-is and as a continuous first-to-last span so standard QA architectures can consume them. This single mechanism replaces the usual translate-and-align machinery, and it is why the method is simple, cheap, and language-agnostic within the service's language coverage.

What would settle it

Manually audit a larger random sample, including passages with overlapping answer spans, long answers, and translations that reorder clauses; if the color-to-span mapping is lost in more than a few percent of overlapping cases, or if the 87.2% correct rate drops sharply on the larger sample, the central claim weakens. A quantitative test would be to have native Finnish speakers re-annotate the same translated passages and compare a QA model trained on the machine-translated spans against one trained on the human re-annotations, holding everything else fixed.

Watch

Extended reading notes

Core claim

The central discovery is that formatted-document transfer through DeepL carries answer-span annotations across languages with high fidelity: when each answer span is rendered in its own color in the source document, the translated document arrives with the coloring intact, and the color-to-span mapping can be reversed mechanically. On SQuAD2.0 to Finnish, 90,233 of 92,749 question-answer pairs survive, a 97.2% preservation rate, and a manual audit of 321 spans finds 87.2% flawless transfers with only 2.2% missing. The resulting Finnish QA models clearly outperform models trained on the earlier Finnish translate-and-align dataset, and a purpose-built Spanish comparison shows the same method beating the alignment-based TAR pipeline.

Load-bearing premise

The load-bearing premise is that DeepL's formatted-document translation preserves the color-coded answer spans reliably and that the colored regions in the output are correct translations of the original answers, a property the paper checks on only 321 manually inspected examples and cannot verify from the vendor.

Editorial extensions

If this is right

The Finnish dataset preserves 90,233 of the original 92,749 question-answer pairs, a 97.2% retention rate that is far above the previous Finnish translate-and-align dataset.
A FinBERT-base reader trained on this dataset reaches EM 68.2 and F1 73.7, clearly above the 55.5 EM and 61.9 F1 reported for the previous Finnish SQuAD with a comparable model.
An English-Finnish-English backtranslation experiment loses 8.4 EM points and 5.1 F1 points over two translation rounds, implying a penalty of roughly 4.2 EM and 2.6 F1 points per round.
The same formatting-transfer method applied to Spanish SQuAD beats the alignment-based TAR dataset when trained and tested in-domain (F1 79.2 vs 73.7), and cross-training on the new dataset improves F1 on both test sets.
A manual error analysis of 321 sampled answers finds 87.2% transferred flawlessly, 9.0% over-extended, and none wrong, indicating the dominant error is minor span over-extension.

Reading between the lines

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

Editorial inference: the same pipeline should be directly applicable to other span-annotated tasks, including named entity recognition and event extraction, where the overlap-handling by extra colors removes the limitation noted in the earlier relation-extraction use of the method.
Editorial inference: because the backtranslation experiment estimates per-round noise, one could pre-screen candidate target languages for translation quality by running the method in reverse and measuring downstream F1, before spending any annotation effort.
Editorial inference: the 2.6% discontinuous answer spans are mostly translation artifacts; rephrasing questions or passages to avoid clause-boundary splits might reduce that rate, though the paper keeps discontinuous spans for compatibility with standard architectures.

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.

Desk Editor's Note

read the letter

Short version: this paper turns a small feature of DeepL—translating formatted documents—into a cheap way to build span-annotated QA datasets in new languages. The trick is not new; the authors already used it for relation extraction in Bassignana et al. (2023), and they cite it. What is new is the QA adaptation with overlapping spans, the resulting Finnish SQuAD2.0, and a genuinely multi-pronged evaluation. That is a solid, reproducible contribution within the subfield.

Credit where due: 87.2% of 321 manually inspected answer spans are fully correct; the backtranslation drops EM by 8.4 points over two rounds, suggesting roughly half that per direction if errors are symmetric; and the Spanish head-to-head against Translate-Align-Retrieve is the strongest evidence, with F1 improving even when training on the DeepL data and testing on TAR data. The dataset also retains 97.2% of the original question-answer pairs, compared with 71.9% for the earlier Finnish SQuAD. Code, data, and a trained model are released openly.

The soft spots are real but not fatal. The headline Finnish comparison (73.7 vs 61.9 F1) is not apples-to-apples: train and dev come from different translation pipelines, and there is no human-annotated Finnish test set or Finnish cross-dataset evaluation. Because the train and dev splits share one DeepL pass, systematic translation errors can be learned and inflate the score relative to performance on naturally occurring Finnish. The manual error analysis checks fidelity to the English source, not naturalness of the Finnish, and samples only answerable questions. The backtranslation estimate leans on an unverified assumption about symmetric error rates. And the method depends on a closed commercial service with no formatting-fidelity guarantee. The authors acknowledge all of these in the Limitations section; they do not hide the dependence, which is more than many dataset papers do.

Net: the resource is worth having, the method is likely to transfer to other languages, and the evaluation is honest. The 'consistently better' claim is directly supported for Spanish and only indirectly for Finnish—a reason to soften the wording, not to reject. I would send this to review, and I would cite the dataset if I were doing Finnish QA.

Referee Report

4 major / 6 minor

Summary. The paper introduces a method for transferring span-level question answering annotations during machine translation by exploiting DeepL's formatted-document translation: answer spans are encoded as colored text in the input document, the translated document preserves the colors, and the spans are reconstructed after translation, with only trailing punctuation stripped. The authors release a Finnish translation of SQuAD2.0 that preserves 90,233 of 92,749 answerable question-answer pairs (97.2%), train extractive QA readers on it (FinBERT, FinALBERT, BERT-large), and evaluate the resource through downstream QA performance on the translated validation set, an English-Finnish-English backtranslation experiment, a Spanish head-to-head comparison with the TAR method, and manual inspection of 321 answer spans. They report that their Finnish dataset gives F1 73.7 with BERT-base versus F1 61.9 for the prior Finnish SQuAD by Kylliäinen, and argue that the method is simple, cheap, and broadly applicable to other span-annotated datasets and languages.

Significance. If the claims hold, the contribution is practically valuable: a low-cost, near-lossless annotation-transfer method for span-labeled datasets, together with an openly released Finnish QA resource and trained model. The paper ships code and data under open licenses, and it reports several evaluation angles rather than a single metric. The manual evaluation (87.2% correct in a 321-example sample), the two-round backtranslation penalty (8.4 EM / 5.1 F1), and the Spanish cross-dataset comparison (Table 3) are meaningful positive evidence. The main caveat is that the headline Finnish quality claim rests on training and evaluating within the same DeepL-translated pipeline, and lacks a shared or manually annotated Finnish test set; consequently the general superiority claim is strongest for Spanish and only indirect for Finnish.

major comments (4)

[Sec. 4.3, Table 1] The claim that the new Finnish data are 'consistently better' than Kylliäinen's Finnish SQuAD is not established by the reported F1 comparison, because the two models are trained and evaluated on different translation pipelines. A model trained on DeepL output can learn systematic DeepL translation patterns present in both train and dev, inflating scores without improving performance on naturally occurring Finnish. Please report cross-evaluation (training on one dataset and evaluating on the other's dev set) or a manually translated Finnish test set; the Spanish experiment in Table 3 demonstrates that this design is feasible and should be replicated for Finnish.
[Sec. 4.4, Table 4] The manual error analysis only samples answerable questions and judges span fidelity relative to the English original; it does not assess the grammaticality or naturalness of the Finnish answer text, nor whether unanswerable questions remain unanswerable after translation. It therefore supports the mechanism of span transfer but not the target-language quality, which is the load-bearing part of the resource claim for Finnish.
[Sec. 4.2, Table 2] The backtranslation experiment estimates the single-round penalty by halving the two-round drop in EM and F1, which relies on the unverified assumption that translation errors are approximately evenly distributed between the en->fi and fi->en rounds. Since Finnish QA is the primary contribution, this assumption should be validated, for example on a small manually translated Finnish test set, or the derived 4.2 EM / 2.6 F1 figures should be presented explicitly as a rough heuristic rather than as the expected one-round loss.
[Sec. 4.3, Table 3] In the Spanish head-to-head, training on the new method's data yields lower EM on the TAR test set (64.5 vs 66.3) while improving F1 (74.0 vs 73.7). The paper notes the 'opposite tendency' of EM but does not investigate it, even though the abstract claims the method produces 'consistently better translated data'. A short analysis of span-boundary statistics, answer lengths, or entity types would help clarify whether the F1 gain reflects overall translation quality rather than a systematic boundary bias.

minor comments (6)

[Sec. 3.3] Please specify the grid-search space and the criterion used to select the FinALBERT pretraining learning rate (5.28e-3), including whether a held-out split was used, so that the negative FinALBERT result can be reproduced and interpreted.
[Sec. 3.2] The paper reports preserving 90,233 of 92,749 answerable QA pairs but does not state how many of the 49,434 unanswerable SQuAD2.0 questions are preserved in the released dataset; please clarify the full dataset statistics.
[Sec. 4.1, Table 1] The cross-dataset comparison in Table 1 mixes model sizes, pretrained backbones, and evaluation splits; adding a column or footnote stating the model type, training split, and eval split for each row would make the comparison more transparent.
[Sec. 3.3] The statement 'the training data used was identical to that used to train the FinBERT model' is too coarse for reproducibility; please report the pretraining steps, batch size, and checkpoint selection procedure for FinALBERT.
[Sec. 3.3] The cost figure 'approximately 20C' should be written as '€20' or '20 EUR', with a note on the conversion rate or pricing tier used.
[Table 1, references] The 'web source' entries in Table 1 lack full publication details; consider moving the URLs into the bibliography or a dedicated data-availability section so that the cited online resources remain traceable.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the Finnish SQuAD evaluation is anchored by external checks (backtranslation to English SQuAD2.0, manual span-fidelity analysis, and Spanish cross-dataset experiments), and no load-bearing claim reduces to a fitted parameter or self-citation.

full rationale

The paper's central claims are that DeepL's formatted-document transfer preserves answer spans and that the resulting Finnish SQuAD is a useful training resource. Neither claim is definitionally tied to its own output. The in-pipeline QA scores (Section 4.1, Table 1) are a standard train/dev evaluation of a released dataset, not a fitted parameter renamed as a prediction; the held-out dev split is produced by the same translation pipeline, which is a generalization/validity limitation but not a circular reduction. The substantive quality claims are independently supported: the backtranslation experiment (Section 4.2, Table 2) evaluates a model trained on backtranslated English against the original English SQuAD2.0 test set, giving an external anchor; the manual error analysis (Section 4.4, Table 4) checks transferred spans against English originals; and the Spanish comparison (Section 4.3, Table 3) includes cross-dataset train/test combinations. The only self-citation, Bassignana et al. (2023) for prior use of DeepL formatting, is transparent and not load-bearing. Comparing F1 with Kylliäinen's Finnish SQuAD on different test sets is a comparability caveat, but the difference is not asserted as a derivation and is supplemented by the external checks. No equation or definition in the paper makes a claimed result equal to its input; therefore no specific circular step can be exhibited.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The central claim rests on DeepL's behavior as a closed service, the assumption that color-marked spans survive translation with correct boundaries, and an even-error assumption in the backtranslation estimate. All are acknowledged in the paper.

free parameters (1)

FinALBERT pretraining learning rate = 5.28e-3
Selected by grid search for the ALBERT pretraining experiment; tangential to the central MT claim.

assumptions (4)

domain assumption DeepL preserves color-coded span formatting and translates the text within each span independently enough to recover answer boundaries
Central mechanism in Section 3.1; no guarantee from the service, validated only indirectly through downstream performance and manual analysis.
domain assumption Backtranslation error is approximately evenly distributed across the two translation rounds
Used in Section 4.2 to estimate single-round MT penalty; explicitly stated as an assumption.
domain assumption Standard SQuAD EM/F1 and BERT fine-tuning are valid for evaluating the translated dataset
Evaluation methodology in Section 4; assumes metric portability to Finnish.
domain assumption Splitting overlapping answer spans into separate colored spans does not degrade translation quality
Design choice in Section 3.1; authors note they observed no clear degradation for this dataset.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Finnish SQuAD: A Simple Approach to Machine Translation of Span Annotations." pith.science (2026). https://pith.science/paper/NCPYKCGY

@misc{pith2026250105963,
  author       = {Pith},
  title        = {Pith review of: Finnish SQuAD: A Simple Approach to Machine Translation of Span Annotations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NCPYKCGY}},
  note         = {Machine review of arXiv:2501.05963}
}

read the original abstract

We apply a simple method to machine translate datasets with span-level annotation using the DeepL MT service and its ability to translate formatted documents. Using this method, we produce a Finnish version of the SQuAD2.0 question answering dataset and train QA retriever models on this new dataset. We evaluate the quality of the dataset and more generally the MT method through direct evaluation, indirect comparison to other similar datasets, a backtranslation experiment, as well as through the performance of downstream trained QA models. In all these evaluations, we find that the method of transfer is not only simple to use but produces consistently better translated data. Given its good performance on the SQuAD dataset, it is likely the method can be used to translate other similar span-annotated datasets for other tasks and languages as well. All code and data is available under an open license: data at HuggingFace TurkuNLP/squad_v2_fi, code on GitHub TurkuNLP/squad2-fi, and model at HuggingFace TurkuNLP/bert-base-finnish-cased-squad2.

Figures

Figures reproduced from arXiv: 2501.05963 by the authors.

**Figure 1.** Observing that oftentimes the answer spans were over-extended by a trailing punctuation symbol during translation, the only post-processing we apply is to strip from each translated span any trailing punctuation. This, in our view, has no negative impact on the QA task. One aspect, common to all machine translation approaches to SQuAD irrespective of the method of annotation transfer, is that the answer spans in t… view at source ↗

**Figure 1.** Example of the colored answer spans from an actual SQuAD passage: the original English [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 12 canonical work pages

[1]

URL: " 'urlintro :=

ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRINGS urlintro eprinturl eprintpr...
[2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...
[3]

Negin Abadani, Jamshid Mozafari, Afsaneh Fatemi, Mohammd Ali Nematbakhsh, and Arefeh Kazemi. 2021. https://doi.org/10.1109/ICWR51868.2021.9443126 ParSQuAD : Machine Translated SQuAD dataset for Persian Question Answering . In 2021 7th International Conference on Web Research ( ICWR ) , pages 163--168, Tehran, Iran. IEEE

work page arXiv 2021
[4]

Elisa Bassignana, Filip Ginter, Sampo Pyysalo, Rob Goot, and Barbara Plank. 2023. https://aclanthology.org/2023.nodalida-1.9 Multi- C ross RE a multi-lingual multi-domain dataset for relation extraction . In Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa), pages 80--85, T \'o rshavn, Faroe Islands. University of Tartu Library

work page 2023
[5]

Casimiro Pio Carrino, Marta R Costa-jussa, and Jose A R Fonollosa. 2019. https://arxiv.org/abs/1912.05200 Automatic Spanish Translation of the SQuAD Dataset for Multilingual Question Answering . ArXiv:1912.05200v2 [cs]

arXiv 2019
[6]

Yiming Cui, Ting Liu, Wanxiang Che, Li Xiao, Zhipeng Chen, Wentao Ma, Shijin Wang, and Guoping Hu. 2019. https://doi.org/10.18653/v1/D19-1600 A Span - Extraction Dataset for Chinese Machine Reading Comprehension . In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural...

arXiv 2019
[7]

Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. http://arxiv.org/abs/1810.04805 BERT : Pre-training of Deep Bidirectional Transformers for Language Understanding . ArXiv:1810.04805 [cs]

arXiv 2019
[8]

Martin d'Hoffschmidt, Wacim Belblidia, Tom Brendlé, Quentin Heinrich, and Maxime Vidal. 2020. http://arxiv.org/abs/2002.06071 FQuAD : French Question Answering Dataset . ArXiv:2002.06071 [cs]

arXiv 2020

Show all 26 references

[9]

Ilmari Kylli \"a inen and Roman Yangarber. 2023. https://aclanthology.org/2023.nodalida-1.53 Question Answering and Question Generation for Finnish . In Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa), pages 529--540, T \'o rshavn, Faroe Islan...

2023
[10]

Ilmari Kylliäinen. 2022. http://urn.fi/URN:NBN:fi:hulib-202206152626 Neural Factoid Question Answering and Question Generation for Finnish . Master's thesis, University of Helsinki

2022
[11]

Zhenzhong Lan, Mingda Chen, Sebastian Goodman, Kevin Gimpel, Piyush Sharma, and Radu Soricut. 2020. http://arxiv.org/abs/1909.11942 ALBERT : A lite BERT for Self -supervised Learning of Language Representations . ArXiv:1909.11942 [cs]

2020 arXiv
[12]

Seungyoung Lim, Myungji Kim, and Jooyoul Lee. 2019. http://arxiv.org/abs/1909.07005 KorQuAD1.0 : Korean QA Dataset for Machine Reading Comprehension . ArXiv:1909.07005 [cs]

2019 arXiv
[13]

Ofri Masad, Kfir Bar, and Amir Cohen. 2023. https://doi.org/10.18653/v1/2023.ijcnlp-main.11 Automatic Translation of Span-Prediction Datasets . In Proceedings of the 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific C...

2023 doi
[14]

Hussein Mozannar, Karl El Hajal, Elie Maamary, and Hazem Hajj. 2019. http://arxiv.org/abs/1906.05394 Neural Arabic Question Answering . ArXiv:1906.05394 [cs]

2019 arXiv
[15]

Koichi Nagatsuka, Clifford Broni-Bediako, and Masayasu Atsumi. 2021. https://aclanthology.org/2021.ranlp-1.112 Pre-training a BERT with Curriculum Learning by Increasing Block-Size of Input Text . In Proceedings of the International Conference on Recent Advances in Natural Lan...

2021
[16]

Susumu Okazawa. 2021. Swedish translation of SQuAD2.0 . https://github.com/susumu2357/SQuAD_v2_sv Last accessed 20 June 2023

2021
[17]

Pranav Rajpurkar, Robin Jia, and Percy Liang. 2018. http://arxiv.org/abs/1806.03822 Know What You Don 't Know : Unanswerable Questions for SQuAD . Number: arXiv:1806.03822 arXiv:1806.03822 [cs]

2018 arXiv
[18]

Pranav Rajpurkar, Jian Zhang, Konstantin Lopyrev, and Percy Liang. 2016. http://arxiv.org/abs/1606.05250 SQuAD : 100,000+ Questions for Machine Comprehension of Text . Number: arXiv:1606.05250 arXiv:1606.05250 [cs]

2016 arXiv
[19]

Tom M Seinen, Jan A Kors, Erik M van Mulligen, and Peter R Rijnbeek. 2024. https://doi.org/10.1093/jamia/ocae159 Annotation-preserving machine translation of English corpora to validate Dutch clinical concept extraction tools . Journal of the American Medical Informatics Assoc...

2024 doi
[20]

Anastassia Shaitarova, Anne G \"o hring, and Martin Volk. 2023. https://aclanthology.org/2023.nodalida-1.22 Machine vs. Human : Exploring Syntax and Lexicon in G erman Translations , with a Spotlight on Anglicisms . In Proceedings of the 24th Nordic Conference on Computational...

2023
[21]

web source. 2021 a . bert-base-spanish-wwm-cased-finetuned-sqac-finetuned-squad2-es. https://huggingface.co/MMG/bert-base-spanish-wwm-cased-finetuned-sqac-finetuned-squad2-es Last accessed 20 June 2023

2021
[22]

web source. 2021 b . Indobert-qa. https://huggingface.co/Rifky/Indobert-QA Last accessed 20 June 2023

2021
[23]

web source. 2021 c . Indonesian squad. https://github.com/Wikidepia/indonesian_datasets/tree/master/question-answering/squad Last accessed 20 June 2023

2021
[24]

Antti Virtanen, Jenna Kanerva, Rami Ilo, Jouni Luoma, Juhani Luotolahti, Tapio Salakoski, Filip Ginter, and Sampo Pyysalo. 2019. http://arxiv.org/abs/1912.07076 Multilingual is not enough: BERT for Finnish . ArXiv:1912.07076 [cs]

2019 arXiv
[25]

Thomas Wolf, Lysandre Debut, Victor Sanh, Julien Chaumond, Clement Delangue, Anthony Moi, Pierric Cistac, Tim Rault, Remi Louf, Morgan Funtowicz, Joe Davison, Sam Shleifer, Patrick Von Platen, Clara Ma, Yacine Jernite, Julien Plu, Canwen Xu, Teven Le Scao, Sylvain Gugger, Mari...

2020 doi
[26]

Robert Östling and Jörg Tiedemann. 2016. https://doi.org/10.1515/pralin-2016-0013 Efficient Word Alignment with Markov Chain Monte Carlo . The Prague Bulletin of Mathematical Linguistics, 106(1):125--146

2016 doi

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [1]

URL: " 'urlintro :=

ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRINGS urlintro eprinturl eprintpr...

[2] [2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

[3] [3]

Negin Abadani, Jamshid Mozafari, Afsaneh Fatemi, Mohammd Ali Nematbakhsh, and Arefeh Kazemi. 2021. https://doi.org/10.1109/ICWR51868.2021.9443126 ParSQuAD : Machine Translated SQuAD dataset for Persian Question Answering . In 2021 7th International Conference on Web Research ( ICWR ) , pages 163--168, Tehran, Iran. IEEE

work page arXiv 2021

[4] [4]

Elisa Bassignana, Filip Ginter, Sampo Pyysalo, Rob Goot, and Barbara Plank. 2023. https://aclanthology.org/2023.nodalida-1.9 Multi- C ross RE a multi-lingual multi-domain dataset for relation extraction . In Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa), pages 80--85, T \'o rshavn, Faroe Islands. University of Tartu Library

work page 2023

[5] [5]

Casimiro Pio Carrino, Marta R Costa-jussa, and Jose A R Fonollosa. 2019. https://arxiv.org/abs/1912.05200 Automatic Spanish Translation of the SQuAD Dataset for Multilingual Question Answering . ArXiv:1912.05200v2 [cs]

arXiv 2019

[6] [6]

Yiming Cui, Ting Liu, Wanxiang Che, Li Xiao, Zhipeng Chen, Wentao Ma, Shijin Wang, and Guoping Hu. 2019. https://doi.org/10.18653/v1/D19-1600 A Span - Extraction Dataset for Chinese Machine Reading Comprehension . In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural...

arXiv 2019

[7] [7]

Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. http://arxiv.org/abs/1810.04805 BERT : Pre-training of Deep Bidirectional Transformers for Language Understanding . ArXiv:1810.04805 [cs]

arXiv 2019

[8] [8]

Martin d'Hoffschmidt, Wacim Belblidia, Tom Brendlé, Quentin Heinrich, and Maxime Vidal. 2020. http://arxiv.org/abs/2002.06071 FQuAD : French Question Answering Dataset . ArXiv:2002.06071 [cs]

arXiv 2020