Pith. sign in

REVIEW 3 major objections 4 minor 42 references

Establishing Strong Baselines for the New Decade: Sequence Tagging, Syntactic and Semantic Parsing with BERT

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read This paper claims that a simple recipe—precomputed averaged BERT token embeddings fed into a plain Bi-LSTM-CRF tagger or a stripped-down biaffine parser—beats previous state-of-the-art systems on 11 of 12 English and Chinese benchmarks.

desk verdict Solid, reproducible BERT baseline paper whose '11 of 12 SOTA' claim is probably right but rests partly on unverified external scores, and the semantic-over-syntactic conclusion overreaches. read the letter →

arxiv 1908.04943 v4 pith:T57TC4KI submitted 2019-08-14 cs.CL

classification cs.CL
keywords BERTembeddingspart-of-speechtaggingdependencyparsingsemanticBi-LSTM-CRFbiaffineattentionstate-of-the-artbaselinesEnglishandChineseNLP
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

Three classic NLP tasks—part-of-speech tagging, syntactic dependency parsing, and semantic dependency parsing—can be improved substantially without inventing new architectures. The paper claims that the decisive ingredient is the input representation: token-level embeddings produced by the pretrained bidirectional transformer BERT, averaged over the token's subwords and frozen before being fed into a simple Bi-LSTM-CRF or biaffine model. Evaluated on 12 English and Chinese datasets, these plain models establish new state-of-the-art scores on 11 of them, with average gains of 2.5 percentage points over previously published systems and a maximum gain of 7.5 points. The paper's analysis adds a qualitative claim: BERT embeddings are more useful for semantic distinctions than for syntactic ones, and they maintain their advantage on longer sentences, which is why semantic parsing benefits most.

What carries the argument

The load-bearing object is the average token-level BERT embedding: for each token split into WordPiece subwords, average the vectors of those subwords from BERT's last layer and treat the result as a static input feature for the downstream encoder. The paper chooses averaging over taking the last subword and over inserting BERT embeddings at a hidden layer, based on small pilot experiments on Prague semantic dependencies; either choice would change the results. This mechanism matters because it separates BERT from the downstream model, allowing BERT to act as a drop-in representation layer while the task model stays simple and fast. The same frozen feature is then concatenated with or substituted for pre-trained word embeddings in a Bi-LSTM-CRF for tagging and in a biaffine parser whose output loss changes from softmax to sigmoid to handle graph-structured semantic dependencies.

What would settle it

Take the best previously published system for each of the 12 datasets and run it on the exact tokenized inputs and official evaluation scripts used in this paper; if that system's score equals or exceeds the paper's BERT score on any of the 11 claimed datasets, the '11 of 12' claim fails for that dataset.

Watch

Extended reading notes

Core claim

We show that the year-2019 state of the art in tagging and parsing does not require task-specific depth: take BERT's last-layer hidden states for a token, average the subword vectors, and feed that fixed vector into a plain recurrent tagger or a biaffine parser. On the Penn Treebank and Penn Chinese Treebank for POS tagging and dependency parsing, and on the SemEval 2015/2016 semantic dependency parsing datasets, this recipe sets new best published results on 11 of 12 benchmarks, outperforming previous systems by 2.5 points on average and by 7.5 points in the largest case. The only dataset where BERT does not surpass the comparison is English POS tagging, where a character-level language-model embedding baseline remains ahead. The paper interprets the pattern as evidence that BERT encodes meaning more than morphology or surface syntax, and that its bidirectional masked-language pretraining gives it an advantage on long-distance dependencies.

Load-bearing premise

The comparison assumes that the previously published scores quoted in Tables 3–6 were produced on the same tokenization, data splits, and evaluation metrics, because those systems are not re-run; a mismatch for any dataset could make a 'new state-of-the-art' result an artifact of the comparison.

Editorial extensions

If this is right

  • Precomputed, averaged BERT embeddings can replace task-specific embedding engineering and still beat published state of the art, so strong baselines do not require GPU-heavy joint fine-tuning of BERT.
  • The same encoder skeleton works for word-level tags, tree-structured dependencies, and graph-structured semantic dependencies in two languages, suggesting the model choice is largely orthogonal to the representation choice.
  • Semantic parsing gains are the largest, up to 7.5 points on Chinese textbook data, which fits the paper's conclusion that BERT embeddings are more semantic than syntactic.
  • On long sentences, BERT-based parsers lose less F1 than recurrent baselines as sentence length grows beyond about 20 tokens, so the embeddings mitigate the distance problem in dependency prediction.
  • The English POS benchmark is the exception where BERT does not beat the character-level baseline, implying that morphological detail from character-level language models is still valuable for morphologically rich tagging.

Reading between the lines

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

  • The frozen-embedding recipe suggests fine-tuning BERT on each downstream task might yield further gains, at the cost of far more compute; the paper does not test this because it deliberately decouples BERT from training.
  • The '11 of 12' count depends on taking previously published scores at face value; an editor would want to re-run those systems on the identical tokenization and evaluation scripts before treating the count as stable.
  • The attention-matrix evidence is observational; a causal test would be to feed BERT embeddings truncated to local context and see whether the long-sentence robustness disappears.
  • If the released code reproduces the tables, the main contribution is a reusable strong baseline rather than a new architecture, which may be the more durable part of the paper.
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 / 4 minor

Summary. The paper applies precomputed BERT token embeddings to three families of structured-prediction models: a Bi-LSTM-CRF tagger for POS tagging, a simplified biaffine parser for dependency parsing, and a biaffine semantic graph parser for semantic dependency parsing. Experiments cover 12 English and Chinese datasets. The authors report that their BERT-based models establish new state-of-the-art results on 11 of 12 datasets, with an average improvement of 2.5 percentage points. They also provide a qualitative analysis of BERT embeddings via self-attention matrices, a sentence-length robustness analysis, and a per-label analysis on one semantic formalism (PSD) leading to the claim that BERT embeddings are more related to semantics than syntax. The paper releases all models and source code and reports three-run averages with standard deviations.

Significance. If the empirical claims hold, the paper makes a valuable contribution: it demonstrates that relatively simple Bi-LSTM-CRF and biaffine models can outperform more complex prior systems on most evaluated benchmarks when supplied with precomputed BERT embeddings. The reproducibility measures are a strength: three independent runs, standard deviations, detailed data splits, hyperparameter tables, and a public code release. The direct comparison against previously published state-of-the-art systems across three tasks and two languages makes the paper a useful reference for future baseline work. The attention-based analysis, while qualitative, offers useful observations about where BERT embeddings help. The main limitation is that the central '11 of 12' claim depends on the comparability of scores reported by external systems that were not re-run, and two claimed wins have very thin margins.

major comments (3)
  1. [§4.2, Table 4(a) and §4.3, Table 5(b)] The central 'new state-of-the-art for 11 of 12 datasets' claim rests on the assumption that the previously published scores in Tables 3–6 were produced under the same data splits, preprocessing, and evaluation metrics as the authors' runs. Since the competing systems were not re-run, the comparison is an audit one. This matters concretely for two of the claimed wins: English dependency parsing LAS 95.29 vs. Clark et al. (2018) 95.0 (margin 0.29), and English out-of-domain PSD 79.48 vs. Dozat and Manning (2018) 79.4 (margin 0.08). These margins are smaller than typical differences introduced by dependency-conversion versions (the paper uses Stanford Parser 3.3.0, per Appendix A.2) or by semantic-scorer variants such as punctuation handling. The authors should either re-run the prior systems under their own pipeline, provide evidence that the external scores are directly comparable (e.g., same conversion and scorer), or explicitly soften the state-of-the-art claim to 'better than previously reported scores on the same official test sets' and identify the two fragile cases.
  2. [§3.4 and Appendix A.2] The biaffine parser input includes a lemma embedding e_LEM_i, but the manuscript does not state whether these lemmas are predicted by the model or taken from gold annotations. If gold lemmas are used, this would give the parser access to information that many prior systems (e.g., Clark et al. 2018, Ma et al. 2018) do not use, making the comparison unfair. The appendix specifies the source of POS tags (jackknifed predictions for English, gold for Chinese) but is silent on lemmas. The authors must state the lemma source; if gold lemmas are used, they should provide a version of the model without gold lemmas or at least discuss the impact.
  3. [§5.3, Abstract, Conclusion] The paper's key finding that 'BERT embeddings are much more related to semantic than syntactic' is drawn entirely from the PSD formalism of the English semantic parsing data, using the difference in per-label F1 scores between baseline and \BERT. This is a single dataset and a single label-set; the conclusion is too broad. The claim appears in the abstract, introduction, and conclusion, so it is a load-bearing part of the paper's narrative. The authors should either restrict the conclusion to PSD (and note that other formalisms or tasks may behave differently) or support the claim with additional evidence, such as the same analysis on DM and PAS, or a comparison of BERT gains on POS versus semantic parsing tasks.
minor comments (4)
  1. [§3.3, Tables 3–6] The notation 'Baseline\ BERT' is confusing; it could be read as set subtraction or as a missing symbol. It is defined in the text, but a clearer label such as 'BERT-only' would improve readability.
  2. [Figure 4] The text immediately preceding Figure 4 contains a garbled line of '/uni00000033/...' characters that appears to be a PDF extraction artifact. This should be removed or corrected in the submitted version.
  3. [Abstract and §4] The abstract states 'outperform the previously best-performing models by 2.5% on average' without specifying how the average is computed across the 12 datasets. Please define the aggregation (e.g., macro-average over all 12 test sets of the difference between the best BERT model and the previous best score).
  4. [Table 14] The claim of model simplicity would be strengthened by reporting parameter counts or inference times alongside the training times, to support the 'simplified' characterization with quantitative evidence.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper's SOTA claims are empirical benchmark results measured against external datasets and published systems, not derived from its own assumptions.

full rationale

This paper is an empirical systems paper: its central claim is that simple taggers and biaffine parsers with precomputed BERT embeddings achieve new state-of-the-art scores on 11 of 12 benchmark datasets. The derivation chain is experimental measurement, not mathematical derivation. The BERT embeddings are an external pretrained resource cited from Devlin et al.; the baseline parsers and taggers are reimplementations of prior architectures; and the comparison scores are taken from published work on the same standard datasets. No fitted parameter is renamed as a prediction, no theory is justified solely by an author self-citation, and no uniqueness or forced-choice argument is imported from the authors' prior work. The only self-citations (e.g., NLP4J for auto-generated POS tags in Appendix A.2, Choi and McCallum for a transition system mentioned in Related Work) are ancillary tooling or background, not load-bearing. The analysis sections are post hoc interpretations of attention weights and sentence-length behavior, not predictions that would reduce by construction. Concerns about whether external published scores were produced under identical preprocessing or evaluation metrics are legitimate correctness or reproducibility risks, but they are not circularity: the paper does not define its SOTA target in terms of its own outputs, nor does it fit a quantity and then predict that same quantity. Therefore no circular step can be exhibited, and the appropriate score is 0.

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

The paper's central claim is empirical rather than a derivation, so the ledger contains no fitted constants or invented entities. The load-bearing assumptions are benchmark comparability and the use of BERT as a frozen external representation.

assumptions (3)
  • domain assumption Standard benchmark datasets and splits are annotated correctly and are comparable across systems, including PTB3, CTB5.1, and SemEval 2015/2016.
    Section 4 and Appendix A define the splits; the state-of-the-art claim compares against published numbers from other systems on these datasets.
  • domain assumption Published previous state-of-the-art scores were obtained under the same preprocessing and evaluation protocols, or differences do not affect rankings.
    Tables 3-6 cite external systems without re-running them. Table 3(b) explicitly marks some Chinese results as character-level and not directly comparable, so the external comparison is not fully controlled.
  • domain assumption The frozen BERT embeddings are treated as fixed inputs and are not task-fine-tuned, so improvements are attributed to BERT plus unchanged task models.
    Section 3.2 precomputes eBERT and feeds it as input; no fine-tuning is performed. This is load-bearing for the paper's interpretation of where BERT helps.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Establishing Strong Baselines for the New Decade: Sequence Tagging, Syntactic and Semantic Parsing with BERT." pith.science (2026). https://pith.science/paper/T57TC4KI

@misc{pith2026190804943,
  author       = {Pith},
  title        = {Pith review of: Establishing Strong Baselines for the New Decade: Sequence Tagging, Syntactic and Semantic Parsing with BERT},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/T57TC4KI}},
  note         = {Machine review of arXiv:1908.04943}
}
read the original abstract

This paper presents new state-of-the-art models for three tasks, part-of-speech tagging, syntactic parsing, and semantic parsing, using the cutting-edge contextualized embedding framework known as BERT. For each task, we first replicate and simplify the current state-of-the-art approach to enhance its model efficiency. We then evaluate our simplified approaches on those three tasks using token embeddings generated by BERT. 12 datasets in both English and Chinese are used for our experiments. The BERT models outperform the previously best-performing models by 2.5% on average (7.5% for the most significant case). Moreover, an in-depth analysis on the impact of BERT embeddings is provided using self-attention, which helps understanding in this rich yet representation. All models and source codes are available in public so that researchers can improve upon and utilize them to establish strong baselines for the next decade.

Figures

Figures reproduced from arXiv: 1908.04943 by the authors.

Figure 1
Figure 1. Generating the Flair embedding for ‘apple’. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Biaffine attention parser Given the i’th token wi , the feature vector is created by concatenating its pre-trained lemma embedding e LEM i , POS embedding e POS i learned during train￾ing and the representation e BERT i from the last layer of BERT. This feature vector is fed into Bi-LSTM, generating two recurrent states r f i and r b i : r f i = LSTMforward e LEM i ⊕ e POS i ⊕ e BERT i  r b i = LSTMbackward e LEM i… view at source ↗
Figure 3
Figure 3. Averaged attention matrices on sentences with 30 tokens. Each cell depicts the attention weight between [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Attention matrices for the Chinese sentence: [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Average labeled F1 scores for semantic parsing w.r.t. sentence lengths. LAS and UAS are represented by [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: ”Share earnings are reported on a fully diluted basis , by company tradition .” [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Top-5 labels that \BERT outperforms the baseline (in red) and vice versa (in blue) on PSD. The baseline performs better on certain arguments involving syntactic relations such as LOC-arg (locative), where the relation usually finds a prepo￾sition as the head of a noun …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 36 canonical work pages

  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]

    Alan Akbik, Duncan Blythe, and Roland Vollgraf. 2018. http://www.aclweb.org/anthology/C18-1139 Contextual String Embeddings for Sequence Labeling . In Proceedings of the 27th International Conference on Computational Linguistics, COLING'18, pages 1638--1649

  4. [4]

    Mariana S. C. Almeida and Andr\' e F. T. Martins. 2015. http://www.aclweb.org/anthology/S15-2162 Lisbon: Evaluating TurboSemanticParser on Multiple Languages and Out-of-Domain Data . In Proceedings of the 9th International Workshop on Semantic Evaluation, SemEval'15, pages 970--973

  5. [5]

    Artsiom Artsymenia, Palina Dounar, and Maria Yermakovich. 2016. http://www.aclweb.org/anthology/S16-1187 IHS-RD-Belarus at SemEval-2016 Task 9: Transition-based Chinese Semantic Dependency Parsing with Online Reordering and Bootstrapping . In Proceedings of the 10th International Workshop on Semantic Evaluation, SemEval'16, pages 1207--1211

  6. [6]

    Bernd Bohnet, Ryan McDonald, Gon \ A alo Sim \ A µes, Daniel Andor, Emily Pitler, and Joshua Maynez. 2018. http://www.aclweb.org/anthology/P18-1246 Morphosyntactic Tagging with a Meta-BiLSTM Model over Context Sensitive Token Encodings . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics, ACL'18, pages 2642--2652

  7. [7]

    Piotr Bojanowski, Edouard Grave, Armand Joulin, and Tomas Mikolov. 2017. https://transacl.org/ojs/index.php/tacl/article/view/999 Enriching Word Vectors with Subword Information . Transactions of the Association for Computational Linguistics, 5:135--146

  8. [8]

    Wanxiang Che, Yanqiu Shao, Ting Liu, and Yu Ding. 2016. http://www.aclweb.org/anthology/S16-1167 SemEval-2016 Task 9: Chinese Semantic Dependency Parsing . In Proceedings of the 10th International Workshop on Semantic Evaluation, SemEval'16, pages 1074--1080

Show all 42 references
  1. [9]

    Do Kook Choe and Eugene Charniak. 2016. https://aclweb.org/anthology/D16-1257 Parsing as Language Modeling . In Proceedings of the Conference on Empirical Methods in Natural Language Processing, EMNLP'16, pages 2331--2336

  2. [10]

    Jinho D. Choi. 2016. http://www.aclweb.org/anthology/N16-1031 Dynamic Feature Induction: The Last Gist to the State-of-the-Art . In Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL'...

  3. [11]

    Choi and Andrew McCallum

    Jinho D. Choi and Andrew McCallum. 2013. http://mirror.aclweb.org/acl2013/ Transition-based Dependency Parsing with Selectional Branching . In Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics, ACL'13, pages 1052--1062

  4. [12]

    Silvie Cinkov \'a , Jan Haji c , Marie Mikulov \'a , Lucie Mladov \'a , Anja Nedolu z ko, Petr Pajas, Jarmila Panevov \'a , Ji r \' Semeck \`y , Jana S indlerov \'a , Josef Toman, et al. 2006. https://ufal.mff.cuni.cz/techrep/tr35.pdf Annotation of English on the Tectogrammati...

  5. [13]

    Manning, and Quoc Le

    Kevin Clark, Minh-Thang Luong, Christopher D. Manning, and Quoc Le. 2018. http://www.aclweb.org/anthology/D18-1217 Semi-Supervised Sequence Modeling with Cross-View Training . In Proceedings of the Conference on Empirical Methods in Natural Language Processing, EMNLP'18, pages...

  6. [14]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2018. http://arxiv.org/abs/1810.04805 BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding . arXiv, 1810.04805

  7. [15]

    Timothy Dozat and Christopher D. Manning. 2017. https://openreview.net/pdf?id=Hk95PK9le Deep Biaffine Attention for Neural Dependency Parsing . In Proceedings of the 5th International Conference on Learning Representations, ICLR'17

  8. [16]

    Timothy Dozat and Christopher D. Manning. 2018. http://www.aclweb.org/anthology/P18-2077 Simpler but More Accurate Semantic Dependency Parsing . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics, ACL'18, pages 484--490

  9. [17]

    Yantao Du, Fan Zhang, Xun Zhang, Weiwei Sun, and Xiaojun Wan. 2015. http://www.aclweb.org/anthology/S15-2154 Peking: Building Semantic Dependency Graphs with a Hybrid Parser . In Proceedings of the 9th International Workshop on Semantic Evaluation, SemEval'15, pages 927--931

  10. [18]

    Jun Hatori, Takuya Matsuzaki, Yusuke Miyao, and Jun'ichi Tsujii. 2011. http://www.aclweb.org/anthology/I11-1136 Incremental Joint POS Tagging and Dependency Parsing in Chinese . In Proceedings of 5th International Joint Conference on Natural Language Processing, IJCNLP'11, pag...

  11. [19]

    Zhiheng Huang, Wei Xu, and Kai Yu. 2015. https://arxiv.org/abs/1508.01991 Bidirectional LSTM-CRF Models for Sequence Tagging . arXiv, 1508.01991

  12. [20]

    Zhongqiang Huang, Mary Harper, and Wen Wang. 2007. http://www.aclweb.org/anthology/D/D07/D07-1117 Mandarin Part-of-Speech Tagging and Discriminative Reranking . In Proceedings of the Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural...

  13. [21]

    Armand Joulin, Edouard Grave, Piotr Bojanowski, and Tomas Mikolov. 2017. https://www.aclweb.org/anthology/E17-2068 Bag of Tricks for Efficient Text Classification . In Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics, ...

  14. [22]

    Urvashi Khandelwal, He He, Peng Qi, and Dan Jurafsky. 2018. http://www.aclweb.org/anthology/P18-1027 Sharp Nearby, Fuzzy Far Away: How Neural Language Models Use Context . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics, ACL'18, pages...

  15. [23]

    Adhiguna Kuncoro, Miguel Ballesteros, Lingpeng Kong, Chris Dyer, Graham Neubig, and Noah A. Smith. 2017. http://www.aclweb.org/anthology/E17-1117 What Do Recurrent Neural Network Grammars Learn About Syntax? In Proceedings of the 15th Conference of the European Chapter of the ...

  16. [24]

    Shuhei Kurita, Daisuke Kawahara, and Sadao Kurohashi. 2017. http://aclweb.org/anthology/P17-1111 Neural Joint Model for Transition-based Chinese Syntactic Analysis . In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, ACL'17, pages 1204--1214

  17. [25]

    Wang Ling, Chris Dyer, Alan W Black, Isabel Trancoso, Ramon Fermandez, Silvio Amir, Luis Marujo, and Tiago Luis. 2015. http://aclweb.org/anthology/D15-1176 Finding Function in Form: Compositional Character Models for Open Vocabulary Word Representation . In Proceedings of the ...

  18. [26]

    Thang Luong, Hieu Pham, and Christopher D. Manning. 2015. http://aclweb.org/anthology/D15-1166 Effective Approaches to Attention-based Neural Machine Translation . In Proceedings of the Conference on Empirical Methods in Natural Language Processing, EMNLP'15, pages 1412--1421

  19. [27]

    Xuezhe Ma and Eduard Hovy. 2016. http://www.aclweb.org/anthology/P16-1101 End-to-end Sequence Labeling via Bi-directional LSTM-CNNs-CRF . In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, ACL'16, pages 1064--1074

  20. [28]

    Xuezhe Ma, Zecong Hu, Jingzhou Liu, Nanyun Peng, Graham Neubig, and Eduard Hovy. 2018. http://www.aclweb.org/anthology/P18-1130 Stack-Pointer Networks for Dependency Parsing . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics, ACL'18, p...

  21. [29]

    Marcus, Mary Ann Marcinkiewicz, and Beatrice Santorini

    Mitchell P. Marcus, Mary Ann Marcinkiewicz, and Beatrice Santorini. 1993. http://aclweb.org/anthology/J93-2004 Building a Large Annotated Corpus of English: The Penn Treebank . Computational Linguistics, 19(2):313--330

  22. [30]

    Tomas Mikolov, Ilya Sutskever, Kai Chen, Greg S Corrado, and Jeff Dean. 2013. https://papers.nips.cc/paper/5021-distributed-representations-of-words-and-phrases-and-their-compositionality.pdf Distributed Representations of Words and Phrases and their Compositionality . In Adva...

  23. [31]

    Stephan Oepen, Marco Kuhlmann, Yusuke Miyao, Daniel Zeman, Silvie Cinkova, Dan Flickinger, Jan Hajic, and Zdenka Uresova. 2015. http://www.aclweb.org/anthology/S15-2153 SemEval 2015 Task 18: Broad-Coverage Semantic Dependency Parsing . In Proceedings of the 9th International W...

  24. [32]

    Hao Peng, Sam Thomson, and Noah A. Smith. 2017. http://aclweb.org/anthology/P17-1186 Deep Multitask Learning for Semantic Dependency Parsing . In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, ACL'17, pages 2037--2048

  25. [33]

    Jeffrey Pennington, Richard Socher, and Christopher Manning. 2014. http://aclweb.org/anthology/D14-1162 GloVe: Global Vectors for Word Representation . In Proceedings of the conference on empirical methods in natural language processing, EMNLP'14, pages 1532--1543

  26. [34]

    Matthew Peters, Mark Neumann, Mohit Iyyer, Matt Gardner, Christopher Clark, Kenton Lee, and Luke Zettlemoyer. 2018. http://www.aclweb.org/anthology/N18-1202 Deep Contextualized Word Representations . In Proceedings of the Conference of the North American Chapter of the Associa...

  27. [35]

    Alec Radford, Karthik Narasimhan, Tim Salimans, and Ilya Sutskever. 2018. https://s3-us-west-2.amazonaws.com/openai-assets/research-covers/language-unsupervised/language_understanding_paper.pdf Improving Language Understanding by Generative Pre-training . Technical report, OpenAI

  28. [36]

    Yuxuan Wang, Wanxiang Che, Jiang Guo, and Ting Liu. 2018. https://aaai.org/ocs/index.php/AAAI/AAAI18/paper/view/16549/16113 A Neural Transition-Based Approach for Semantic Dependency Graph Parsing . In Proceedings of the 32nd AAAI Conference on Artificial Intelligence, AAAI'18...

  29. [37]

    Zhiguo Wang and Nianwen Xue. 2014. http://www.aclweb.org/anthology/P14-1069 Joint POS Tagging and Transition-based Constituent Parsing in Chinese with Non-local Features . In Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics, ACL'14, pages...

  30. [38]

    Yonghui Wu, Mike Schuster, Zhifeng Chen, Quoc V Le, Mohammad Norouzi, Wolfgang Macherey, Maxim Krikun, Yuan Cao, Qin Gao, Klaus Macherey, et al. 2016. https://arxiv.org/abs/1609.08144 Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Transl...

  31. [39]

    Naiwen Xue, Fei Xia, Fu-dong Chiou, and Marta Palmer. 2005. https://doi.org/10.1017/S135132490400364X The Penn Chinese TreeBank: Phrase Structure Annotation of a Large Corpus . Natural Language Engineering, 11(2):207--238

  32. [40]

    Meishan Zhang, Yue Zhang, Wanxiang Che, and Ting Liu. 2014. http://www.aclweb.org/anthology/P14-1125 Character-Level Chinese Dependency Parsing . In Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics, ACL'14, pages 1326--1336

  33. [41]

    Yuan Zhang, Chengtao Li, Regina Barzilay, and Kareem Darwish. 2015. http://www.aclweb.org/anthology/N15-1005 Randomized Greedy Inference for Joint Segmentation, POS Tagging and Dependency Parsing . In Proceedings of the Conference of the North American Chapter of the Associati...

  34. [42]

    Yue Zhang and Stephen Clark. 2008. http://www.aclweb.org/anthology/D08-1059 A Tale of Two Parsers: Investigating and Combining Graph-based and Transition-based Dependency Parsing . In Proceedings of the Conference on Empirical Methods in Natural Language Processing, EMNLP'08, ...

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.