Pith. sign in

REVIEW 3 major objections 5 minor 37 references

Summarisation of German Judgments in conjunction with a Class-based Evaluation

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Marking legal entities in German judgments before fine-tuning helps a decoder-only language model find the content that belongs in guiding principles, but the generated summaries still need expert revision before real-world use.

desk verdict A useful seven-class evaluation framework for legal summaries, attached to an entity-enrichment claim that needs more than single training runs to believe. read the letter →

arxiv 2505.05947 v1 pith:S6KEHHPV submitted 2025-05-09 cs.CL

Bianca Steffes , Nils Torben Wiedemann , Alexander Gratz , Pamela Hochreither , Jana Elina Meyer , Katharina Luise Schilke This is my paper · ORCID
classification cs.CL
keywords legalsummarisationguidingprinciplesGermanjudgmentsentityrecognitiondecoder-onlylanguagemodelevaluationclassesROUGEBERTScore
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

German higher courts routinely publish guiding principles, short headnotes stating the essence of a judgment. This paper tries to generate those headnotes automatically for decisions of the German Federal Court of Justice by fine-tuning a decoder-based 7B language model, once on plain judgments and once on judgments whose legal entities, such as statutes, court decisions, legal literature, and case-specific rules, are wrapped in tags. The paper argues that the entity-enriched model finds the relevant content better: it scores slightly higher on ROUGE and BERTScore and, in a manual review by legal experts, satisfies the main-focus class for 42% of judgments versus 32% for the plain model. The paper also proposes a seven-class evaluation scheme covering intelligibility, language, pertinence, completeness, main-focus completeness, correctness, and superiority, and shows that legal experts apply it with substantial overall agreement. Its own conclusion is that the summaries are not yet good enough for practice, mainly because many contain hallucinated citations or imprecise legal wording.

What carries the argument

The load-bearing mechanism is a two-stage pipeline. First, a legal-entity recogniser trained for German legal documents inserts entity tags into the reasons-for-decision sections used as training input, wrapping references such as § 125 BGB as <GS> § 125 BGB </GS>, and these tags are added as special tokens to the model vocabulary. Second, a decoder-only 7B language model is fine-tuned on the tagged texts to generate guiding principles. The evaluation machinery is the proposed seven-class rubric, applied by five legal professionals with three reviewers per judgment and a two-of-three majority rule; it converts the abstract qualities of language, pertinence, completeness, and correctness into countable fulfilment rates. The entity tags are the hypothesized salience signal that helps the model identify which content the headnote should carry.

What would settle it

Train the same model on the same data with several random seeds for both the plain and entity-tagged conditions and compare the distributions of class 5 fulfilment and ROUGE on the test set; if the tagged condition does not beat the plain condition in the large majority of seed pairs, the claim that legal entities help content selection is not supported. A cheaper second check is to test whether the extra aspects captured by the tagged model in class 5 judgments are actually mentioned in the entity-tagged portions of the source judgment.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that enriching training texts with legal-entity tags improves the generative model's content selection without fixing its output quality. The model fine-tuned on tagged judgments (LerLeoLM) outperforms the model fine-tuned on plain text (LeoLM) and the LexRank extractive baseline across all automated metrics and in the average number of evaluation classes fulfilled; the clearest manual signal is class 5, main focus, where the tagged model reaches 42% fulfilment versus 32% for the plain model and 17% for the baseline. The same manual evaluation shows only about 10% of generated summaries are complete (class 4), about 20 to 23% are pertinent (class 3), and reviewers report invented citations, so the authors state that the quality is insufficient for practical use without expert revision.

Load-bearing premise

The main positive result depends on the assumption that the difference between the two trained models, one with legal-entity tags and one without, is a real effect of the tags rather than random variation; the comparison is based on 60 manually rated judgments, one training run per condition, and no statistical significance test.

Editorial extensions

If this is right

  • If the entity-enrichment effect is real, any legal summarisation system can adopt tagging as a cheap preprocessing step that improves content selection with no change to architecture.
  • The class-based evaluation reveals gains that ROUGE and BERTScore understate: the class 5 gap of 0.42 versus 0.32 is larger than most metric differences, so richer manual rubrics are worth the cost when the use case is legal.
  • The pattern that short single-area guiding principles summarise well while mixed procedural-and-substantive-law principles fail suggests that practical systems should predict difficulty and escalate to human review.
  • Because hallucinated citations persist even in the entity-enriched model, production deployment of such summarisers requires a fact-checking or citation-verification step.
  • The seven classes, or a subset of them, can be transferred to other languages and legal document types whose summaries need pertinence and correctness checks rather than surface overlap.

Reading between the lines

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

  • A natural next experiment the paper does not run is to vary the entity types in the tags, for example norms versus parties versus court decisions, to see which type drives the class 5 gain; the paper's data permit this by retagging with one entity type at a time.
  • Because the paper trains only one model per condition, the headline 0.42-versus-0.32 gap could be partly seed noise; repeating the fine-tuning with several seeds would turn the observed gain into an interval estimate.
  • The low reviewer agreement on classes 2 and 3 suggests the rubric needs anchor examples; a version with one worked example per class would likely sharpen the instrument more than adding another judge.
  • If the entity-tagging benefit transfers, it offers a direct replication design for other low-resource legal languages: tag a small corpus with any available named-entity tool and measure class 5 fulfilment before and after.
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

3 major / 5 minor

Summary. The paper addresses the automated summarisation of German Federal Court of Justice (BGH) judgments into so-called guiding principles. The authors fine-tune a decoder-based large language model (LeoLM-Mistral-7B) on a corpus of 5,081 judgments, once on the original reasons-for-decision text and once on text enriched with legal-entity tags derived from a German legal named-entity recognizer. They also propose a seven-class manual evaluation scheme (intelligibility, language, pertinence, completeness, main focus, correctness, superiority) built on publishing guidelines from C.H. Beck. The models are compared against each other and against an extractive LexRank baseline using ROUGE, BERTScore, and the proposed classes, applied by five legal experts among the authors. The paper's central claim, stated in the abstract, is that legal-entity enrichment helps the generative model find relevant content, while also acknowledging that the generated summaries are not yet suitable for practical use.

Significance. If the central claim is sound, the work makes two useful contributions: (i) evidence that entity-level annotation of training data can improve content selection for abstractive legal summarisation in German, a language with relatively little prior work, and (ii) a multi-faceted manual evaluation instrument for legal summaries that goes beyond ROUGE/BERTScore. The paper is transparent and careful in several respects: it releases its code, reports the data construction in detail, discusses reviewer agreement honestly, and explicitly reports the negative finding that summaries are not practice-ready. The proposed class-based evaluation is a sensible attempt to capture dimensions that automated metrics miss. However, the positive causal claim about legal entities rests on a thin statistical basis, as detailed in the major comments, and the manual evaluation is performed by the authors who designed both the method and the evaluation scheme.

major comments (3)
  1. [Section 5.1 and Table 8] The central claim that 'employing legal entities helps the generative model to find the relevant content' is supported only by a single fine-tuned run per condition, with no multiple seeds, no confidence intervals, and no significance tests. The headline class-5 difference (LerLeoLM 0.42 versus LeoLM 0.32) is computed on 60 manually rated judgments under a 2-of-3 reviewer majority rule, so it corresponds to roughly six judgments changing their class membership; a 95% confidence interval for this difference would very plausibly include zero. The automated-metric deltas in Table 6 (e.g., ROUGE-1 0.3052 vs. 0.2997; BERTScore 0.6746 vs. 0.6724) are far smaller than the reported standard deviations (around 0.066 for BERTScore). The authors should report results over multiple training seeds, provide per-instance outcomes, and apply a paired significance test (e.g., McNemar for class 5 or a bootstrap for the mean differences). Until this is done, the abstract's first result clause is not statistically established.
  2. [Sections 4.2 and 5.2] The manual evaluation that produces the class-5 result is conducted by the five authors themselves (the second to sixth authors), the same people who designed the seven evaluation classes and who have a stake in the entity-enrichment method. Although the reviews are blinded as to which system produced which summary, the lack of independent assessors creates a risk of expectation effects, especially for class 5, whose inter-rater agreement is only moderate (Fleiss' kappa 0.43 in Table 5). This risk is compounded by the small sample size. The authors should either recruit independent legal experts to re-evaluate at least a subset of the 60 judgments, or report per-reviewer class-5 decisions and agreement statistics specifically for the 60-judgment subset used in Section 5, so that readers can judge whether the 0.10 gap is consistent across raters.
  3. [Section 6, correlations] The paper uses correlations between the manual classes and ROUGE/BERTScore to argue that the classes measure what the metrics intend to measure, but the analysis is reported only via qualitative Cohen labels ('low', 'medium') without correlation coefficients, significance levels, or confidence intervals. Since class 5 is the main outcome supporting the paper's positive claim, the authors should report the actual Spearman coefficients and their uncertainty, or explicitly label the analysis as exploratory and not necessary for the paper's conclusions.
minor comments (5)
  1. [Section 3.1, footnote 4] The footnote contains the typo 'less then 1%' and should read 'less than 1%'.
  2. [Section 5.1] The hyperparameters are given as 'learning rate of 2−4', which is ambiguous; it should be written as '2e-4' (or '2×10⁻⁴') to be clear.
  3. [Section 5.1] The sentence 'we have a model trained on a a single (summari-sation) task' contains a duplicated article 'a' and a hyphenation artifact; please correct.
  4. [Section 6] The sentence 'in it's training' should be 'in its training'.
  5. [Table 5 caption] The caption says 'number of judgments which (not) fulfilled the classes', but the counts (e.g., 591+9=600) correspond to summaries (100 judgments × two approaches × three reviewers), not judgments; please rephrase the caption to refer to summary-reviewer decisions.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the entity-enrichment comparison is an empirical A/B test scored against external gold summaries, not a construction that reduces to its inputs.

full rationale

The paper's central claim is that legal-entity enrichment helps the generative model find relevant content. The derivation chain is an empirical comparison: LeoLM and LerLeoLM are fine-tuned with the same hyperparameters, the only difference being inserted legal-entity tags from an external LER model (Leitner et al., bert-german-ler), and both are evaluated against court-authored gold guiding principles. Automated metrics (ROUGE, BERTScore) and the seven-class manual scheme are independent of the entity-enrichment manipulation. The evaluation classes are anchored in the external C.H. Beck Redaktionsrichtlinie [3], not derived from the model outputs or from the entity tags, so the class-5 gain (0.42 vs. 0.32) is an empirical measurement rather than a forced relation. The gold summaries come from the court, not from the model or the authors' fitting procedure. The single self-citation ([31], Steffes and Rataj) appears only in the related-work review and is not load-bearing. Concerns about a single training run, the absence of significance tests, and the small 60-judgment manual sample are statistical robustness issues, not circularity; they do not make the derivation equivalent to its inputs. No uniqueness theorem, ansatz-smuggling citation, or redefinition of a known result as a prediction is present. The paper even explicitly reports limitations and the need for further validation, which further supports a non-circular presentation.

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

The evaluation is not a derivation; the central claim is empirical. The most load-bearing choices are the gold-standard assumption (original guiding principles), the hand-set evaluation thresholds, and the use of author reviewers. No new physical or conceptual entities are introduced.

free parameters (4)
  • Class 5 completeness threshold = 3/4 (75%) of legal aspects
    Class 5 is the main evidence for the claim that legal entity enrichment helps the model find relevant content; the threshold is chosen by the authors, not derived from any objective standard.
  • Reviewer majority threshold = 2 of 3 reviewers
    Table 8 percentages and the conclusion that LerLeoLM outperforms LeoLM in class 5 depend on counting a class as fulfilled when at least two of three reviewers agree; this is a hand-set aggregation rule.
  • Generation max tokens = 750
    Limits output length for all generated summaries and affects content coverage and comparability; chosen by hand rather than by a data-driven criterion.
  • Training hyperparameters = learning rate 2^-4, 10 epochs, batch size 1
    A single configuration is used without tuning or multiple seeds; the absence of repeated runs means the reported differences between LeoLM and LerLeoLM have no error bars.
assumptions (4)
  • domain assumption The original guiding principles published with BGH judgments are a valid gold standard for summary quality.
    The entire evaluation pipeline, ROUGE/BERTScore and classes 3-5, compares generated summaries against the court's original guiding principles; if the gold summaries themselves are not the right target, the evaluation loses meaning. This is assumed throughout Sections 3.2, 4, and 5.2.
  • domain assumption The C.H. Beck Redaktionsrichtlinie for unofficial guiding principles can be generalized into the seven evaluation classes.
    The classes are built on the C.H. Beck guidelines [3], assuming these publisher guidelines are a sound basis for assessing all BGH guiding principles, including official ones.
  • domain assumption The bert-german-ler model reliably identifies legal entities in the judgments.
    The enrichment step in Section 5.1 relies on this pre-trained NER model from Leitner et al. [18]; if the entity annotations are noisy, the enrichment effect is confounded.
  • domain assumption Reviewer judgments by the author team are unbiased despite authors being the reviewers and knowing the summaries come from an LLM or a baseline.
    The five legal professionals are also authors; they are blinded to which approach created which summary but not to the fact that generative models and LexRank are involved. Bias cannot be ruled out, and this is acknowledged only partially in Section 6.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Summarisation of German Judgments in conjunction with a Class-based Evaluation." pith.science (2026). https://pith.science/paper/S6KEHHPV

@misc{pith2026250505947,
  author       = {Pith},
  title        = {Pith review of: Summarisation of German Judgments in conjunction with a Class-based Evaluation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S6KEHHPV}},
  note         = {Machine review of arXiv:2505.05947}
}
read the original abstract

The automated summarisation of long legal documents can be a great aid for legal experts in their daily work. We automatically create summaries (guiding principles) of German judgments by fine-tuning a decoder-based large language model. We enrich the judgments with information about legal entities before the training. For the evaluation of the created summaries, we define a set of evaluation classes which allows us to measure their language, pertinence, completeness and correctness. Our results show that employing legal entities helps the generative model to find the relevant content, but the quality of the created summaries is not yet sufficient for a use in practice.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 21 canonical work pages

  1. [1]

    Dennis Aumiller, Jing Fan, and Michael Gertz. 2023. On the State of German (Abstractive) Text Summarization. In BTW 2023. Gesellschaft für Informatik e.V., Bonn, 195–220. https://doi.org/10.18420/BTW2023-10

    work page doi:10.18420/btw2023-10 2023

  2. [2]

    Satanjeev Banerjee and Alon Lavie. 2005. METEOR: An Automatic Metric for MT Evaluation with Improved Correlation with Human Judgments. In Proceedings of the ACL Workshop on Intrinsic and Extrinsic Evaluation Measures for Machine Translation and/or Summarization , Jade Goldstein, Alon Lavie, Chin-Yew Lin, and Clare Voss (Eds.). Association for Computationa...

    work page 2005

  3. [3]

    C.H. Beck. 2018. Redaktionsrichtlinie des Verlags C.H.BECK/Franz Vahlen für die Gestaltung von Leitsätzen. https://rsw.beck.de/docs/librariesprovider27/default- document-library/redrl_chbeck_leits%C3%A4tze_1_7_2018.pdf?sfvrsn= b783470d_8 Version dated to 01.07.2018

    work page 2018

  4. [5]

    Jacob Cohen. 1988. Statistical Power Analysis for the Behavioral Sciences . Rout- ledge. https://doi.org/10.4324/9780203771587

    doi:10.4324/9780203771587 1988

  5. [6]

    Tim Dettmers, Artidoro Pagnoni, Ari Holtzman, and Luke Zettlemoyer. 2023. QLoRA: Efficient Finetuning of Quantized LLMs. arXiv preprint arXiv:2305.14314 (2023)

    arXiv 2023

  6. [7]

    Nicholas Egan, Oleg Vasilyev, and John Bohannon. 2022. Play the Shannon Game with Language Models: A Human-Free Approach to Summary Evaluation. Proceedings of the AAAI Conference on Artificial Intelligence 36, 10 (Jun. 2022), 10599–10607. https://doi.org/10.1609/aaai.v36i10.21304

    work page doi:10.1609/aaai.v36i10.21304 2022

  7. [8]

    Mohamed Elaraby, Huihui Xu, Morgan Gray, Kevin Ashley, and Diane Lit- man. 2024. Adding Argumentation into Human Evaluation of Long Document Abstractive Summarization: A Case Study on Legal Opinions. In Proceedings of the Fourth Workshop on Human Evaluation of NLP Systems (HumEval) @ LREC-COLING 2024, Simone Balloccu, Anya Belz, Rudali Huidrom, Ehud Reite...

    work page 2024

  8. [9]

    Ahmed Elnaggar, Christoph Gebendorfer, Ingo Glaser, and Florian Matthes. 2018. Multi-Task Deep Learning for Legal Document Translation, Summarization and Multi-Label Classification. In Proceedings of the 2018 Artificial Intelligence and Cloud Computing Conference (Tokyo, Japan) (AICCC ’18). Association for Computing Machinery, New York, NY, USA, 9–15. htt...

    work page arXiv 2018

Show all 37 references
  1. [10]

    Günes Erkan and Dragomir R. Radev. 2004. LexRank: Graph-based Lexical Centrality as Salience in Text Summarization. Journal of Artificial Intelligence Research 22 (Dec. 2004), 457–479. https://doi.org/10.1613/jair.1523

    2004 doi

  2. [11]

    Atefeh Farzindar and Guy Lapalme. 2004. LetSum, an automatic Legal Text Summarizing system. In Legal Knowledge and Information Systems, Jurix 2004: The Seventeenth Annual Conference. 11–18

    2004

  3. [12]

    Ingo Glaser, Sebastian Moser, and Florian Matthes. 2021. Summarization of Ger- man Court Rulings. InProceedings of the Natural Legal Language Processing Work- shop 2021, Nikolaos Aletras, Ion Androutsopoulos, Leslie Barrett, Catalina Goanta, and Daniel Preotiuc-Pietro (Eds.). ...

    2021 doi

  4. [13]

    Ben Hachey and Claire Grover. 2006. Extractive summarisation of legal texts. Artificial Intelligence and Law 14 (Dec. 2006), 305–345. https://doi.org/10.1007/ s10506-007-9039-z

    2006

  5. [14]

    Zhanying He, Chun Chen, Jiajun Bu, Can Wang, Lijun Zhang, Deng Cai, and Xiaofei He. 2012. Document summarization based on data reconstruction. In Proceedings of the Twenty-Sixth AAAI Conference on Artificial Intelligence(Toronto, Ontario, Canada) (AAAI’12). AAAI Press, 620–626

    2012

  6. [15]

    Mi-Young Kim, Ying Xu, and Randy Goebel. 2013. Summarization of Legal Texts with High Cohesion and Automatic Compression Rate. InNew Frontiers in Artificial Intelligence, Yoichi Motomura, Alastair Butler, and Daisuke Bekki (Eds.). Springer Berlin Heidelberg, Berlin, Heidelberg...

    2013

  7. [16]

    Gerhard Koebler. 2024. Juristisches Wörterbuch (19 ed.). Verlag Franz Vahlen

    2024

  8. [17]

    Richard Landis and Gary G

    J. Richard Landis and Gary G. Koch. 1977. The Measurement of Observer Agreement for Categorical Data. Biometrics 33, 1 (1977), 159–174. http: //www.jstor.org/stable/2529310

    1977

  9. [18]

    Elena Leitner, Georg Rehm, and Julián Moreno-Schneider. 2020. A Dataset of German Legal Documents for Named Entity Recognition. https://doi.org/10. 48550/ARXIV.2003.13016

    2020 doi arXiv

  10. [19]

    Chin-Yew Lin. 2004. ROUGE: A Package for Automatic Evaluation of Summaries. In Text Summarization Branches Out. Association for Computational Linguistics, Barcelona, Spain, 74–81. https://aclanthology.org/W04-1013

    2004

  11. [21]

    Yang Liu and Mirella Lapata. 2019. Text Summarization with Pretrained Encoders. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Pro- cessing (EMNLP-IJCNLP), Kentaro Inui, J...

    2019 doi

  12. [22]

    Hans P. Luhn. 1958. The Automatic Creation of Literature Abstracts.IBM Journal of Research and Development 2, 2 (1958), 159–165. https://doi.org/10.1147/rd.22. 0159

    1958 doi

  13. [23]

    Sewon Min, Kalpesh Krishna, Xinxi Lyu, Mike Lewis, Wen-tau Yih, Pang Koh, Mohit Iyyer, Luke Zettlemoyer, and Hannaneh Hajishirzi. 2023. FActScore: Fine- grained Atomic Evaluation of Factual Precision in Long Form Text Generation. In Proceedings of the 2023 Conference on Empiri...

    2023

  14. [24]

    Moawad and Mostafa Aref

    Ibrahim F. Moawad and Mostafa Aref. 2012. Semantic graph reduction approach for abstractive Text Summarization. In 2012 Seventh International Conference on Computer Engineering & Systems (ICCES) . 132–138. https://doi.org/10.1109/ ICCES.2012.6408498

    2012

  15. [25]

    Gianluca Moro, Nicola Piscaglia, Luca Ragazzi, and Paolo Italiani. 2024. Multi- language transfer learning for low-resource legal case summarization. Artificial Intelligence and Law 32 (Dec. 2024), 1111–1139. https://doi.org/10.1007/s10506- 023-09373-8

    2024 doi

  16. [26]

    Ramesh Nallapati, Feifei Zhai, and Bowen Zhou. 2017. SummaRuNNer: A Re- current Neural Network Based Sequence Model for Extractive Summarization of Documents. In Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence (San Francisco, California, USA) (AAAI’1...

    2017

  17. [27]

    Huyen Nguyen and Junhua Ding. 2023. Keyword-based Augmentation Method to Enhance Abstractive Summarization for Legal Documents. In Proceedings of the Nineteenth International Conference on Artificial Intelligence and Law (ICAIL ’23). Association for Computing Machinery, New Yo...

    2023

  18. [28]

    Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. BLEU: a method for automatic evaluation of machine translation. In Proceedings of the 40th Annual Meeting on Association for Computational Linguistics (Philadel- phia, Pennsylvania) (ACL ’02). Association for C...

    2002

  19. [29]

    Seth Polsley, Pooja Jhunjhunwala, and Ruihong Huang. 2016. CaseSummarizer: A System for Automated Summarization of Legal Texts. In Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: System Demonstrations. The COLING 2016 Organizing Comm...

    2016

  20. [30]

    2024.German Summarization with Large Language Models

    Roy Schubiger. 2024.German Summarization with Large Language Models. Master Thesis. ETH Zurich, Zurich. https://doi.org/10.3929/ethz-b-000659231

    2024 doi

  21. [31]

    Bianca Steffes and Piotr Rataj. 2022. Legal Text Summarization Using Argumen- tative Structures. In Legal Knowledge and Information Systems - Proceedings of the 35th International Conference on Legal Knowledge and Information Systems – JURIX 2022 . IOS Press BV, 243–248. https...

    2022

  22. [32]

    HuiHui Xu and Kevin Ashley. 2023. A Question-Answering Approach to Evalu- ating Legal Summaries. In Legal Knowledge and Information Systems, Jurix 2023 . 293–298

    2023

  23. [33]

    Huihui Xu, Jaromir Savelka, and Kevin D. Ashley. 2021. Toward summarizing case decisions via extracting argument issues, reasons, and conclusions. InProceedings of the Eighteenth International Conference on Artificial Intelligence and Law (São Paulo, Brazil) (ICAIL ’21). Assoc...

    2021

  24. [34]

    Jen-Yuan Yeh, Hao-Ren Ke, Wei-Pang Yang, and I-Heng Meng. 2005. Text summa- rization using a trainable summarizer and latent semantic analysis. Information Processing & Management 41, 1 (2005), 75–95. https://doi.org/10.1016/j.ipm.2004. 04.003 An Asian Digital Libraries Perspective

    2005 doi

  25. [35]

    Jiyoung Yoon, Muhammad Junaid, Sajid Ali, and Jongwuk Lee. 2022. Abstrac- tive Summarization of Korean Legal Cases using Pre-trained Language Models. In 2022 16th International Conference on Ubiquitous Information Management and Communication (IMCOM). 1–7. https://doi.org/10.1...

    2022

  26. [36]

    Weizhe Yuan, Graham Neubig, and Pengfei Liu. 2021. BARTScore: Evaluating Generated Text as Text Generation. arXiv:2106.11520 [cs.CL] https://arxiv.org/ abs/2106.11520

    2021 arXiv

  27. [37]

    Weinberger, and Yoav Artzi

    Tianyi Zhang*, Varsha Kishore*, Felix Wu*, Kilian Q. Weinberger, and Yoav Artzi. 2020. BERTScore: Evaluating Text Generation with BERT. In Interna- tional Conference on Learning Representations . https://openreview.net/forum? id=SkeHuCVFDr

    2020

  28. [38]

    Ashley, and Matthias Grabmair

    Linwu Zhong, Ziyi Zhong, Zinian Zhao, Siyuan Wang, Kevin D. Ashley, and Matthias Grabmair. 2019. Automatic Summarization of Legal Decisions Using Iterative Masking of Predictive Sentences. In Proceedings of the Seventeenth International Conference on Artificial Intelligence an...

    2019

  29. [262]

    https://aclanthology.org/C16-2054

Pith tools

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