Compositional Semantic Parsing Across Graphbanks

Alexander Koller; Jonas Groschwitz; Matthias Lindemann

arxiv: 1906.11746 · v2 · pith:QQL54IYUnew · submitted 2019-06-27 · 💻 cs.CL

Compositional Semantic Parsing Across Graphbanks

Matthias Lindemann , Jonas Groschwitz , Alexander Koller This is my paper

Pith reviewed 2026-05-25 14:53 UTC · model grok-4.3

classification 💻 cs.CL

keywords semantic parsinggraphbankscompositional parserneural networksmulti-task learningBERTAMRDM

0 comments

The pith

A single compositional neural semantic parser reaches competitive accuracy across multiple graphbanks without per-bank hand design.

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

The paper demonstrates that one neural model can map sentences to graph-based meaning representations drawn from several different graphbanks at once. Earlier parsers were built by hand for each individual graphbank, limiting reuse. The new parser uses a compositional architecture to deliver competitive results on a range of these banks, and further gains come from BERT embeddings plus multi-task learning, producing new state-of-the-art numbers on DM, PAS, PSD, AMR 2015 and EDS. If correct, the result shows that general semantic parsers can replace collections of specialized ones. Readers would care because this reduces the need to redesign parsers whenever a new meaning representation format appears.

Core claim

The paper presents a compositional neural semantic parser that, for the first time, attains competitive accuracies across a diverse collection of graphbanks rather than requiring separate hand-designed systems for each bank.

What carries the argument

The compositional neural semantic parser, which assembles graph meanings from sentence parts through neural composition steps.

Load-bearing premise

The parser's cross-graphbank performance comes from genuine compositional generality rather than from tuning that happens to fit the particular set of banks used in testing.

What would settle it

Run the parser on a graphbank withheld from both single-task and multi-task training and check whether accuracy remains competitive with bank-specific systems.

Figures

Figures reproduced from arXiv: 1906.11746 by Alexander Koller, Jonas Groschwitz, Matthias Lindemann.

**Figure 1.** Figure 1: AMR for The shy cat wants to eat with its AM analysis. special root source, drawn here with a bold outline. There are two operations in the AM Algebra that combine as-graphs. First, the apply operation APPX for a source X, as in APPO (Gwant, Geat) with the result shown in Fig. 1e. The operation combines two as-graphs, a head and an argument, by filling the head’s X-source with the root of the argument. No… view at source ↗

**Figure 3.** Figure 3: AMR (a) and PAS (b) for Giraffes are tall. Alignments are given in EDS and not necessary in DM, PAS, and PSD. Grouping. We follow two principles in grouping edges with nodes: Edges between heads and arguments always belong with the head, and edges between heads and modifiers with the modifier (regardless of the direction into which the edge points). This yields supertags that generalize well, e.g. a noun… view at source ↗

**Figure 2.** Figure 2: Semantic representations for The shy cat wants to eat, each with an AM dependency tree below. Similarly, all graphs have an edge indicating that “shy” modifies “cat”, although edge label and edge direction vary. However, the graphbanks differ not only in edge directions and labels, but also structurally. For example, DM, PAS and EDS annotate determiners while AMR and PSD do not [PITH_FULL_IMAGE:figures/f… view at source ↗

**Figure 4.** Figure 4: Coordination in (a) DM and (b, c) PSD. nomena that cause reentrancies in the new graphbanks, but not in AMR – such as copula in PAS, c.f [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗

read the original abstract

Most semantic parsers that map sentences to graph-based meaning representations are hand-designed for specific graphbanks. We present a compositional neural semantic parser which achieves, for the first time, competitive accuracies across a diverse range of graphbanks. Incorporating BERT embeddings and multi-task learning improves the accuracy further, setting new states of the art on DM, PAS, PSD, AMR 2015 and EDS.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

A single model reaches competitive cross-graphbank results, but the abstract leaves open whether compositionality or just multi-task plus BERT is doing the work.

read the letter

The main point is that this paper gets one model to perform competitively on a bunch of different graphbanks for semantic parsing. They have a compositional neural parser that works across DM, PAS, PSD, AMR 2015 and EDS. With BERT and multi-task learning it reaches new state of the art numbers on those. What it does well is showing that you can share training across these representations and get good results without designing separate parsers. The use of BERT is a solid modern addition that boosts performance. The potential issue is whether the compositionality is actually the thing enabling the cross-graphbank success. The stress-test note is on point here: if the model is mostly a standard setup plus multi-task, then the compositional label might not be doing the heavy lifting. Without seeing ablations that test that specifically, it's hard to know how much credit the compositional design deserves. The abstract alone doesn't resolve this, so the full paper needs to have those details. This paper is for NLP folks focused on semantic parsing who want more general systems. A reader looking for ways to reduce hand-engineering would get something out of the results. It should go to peer review. The claim is practically useful if it holds, and the experiments span enough graphbanks to be interesting. Even with possible needs for more analysis on the mechanism, it's worth a referee's time.

Referee Report

3 major / 2 minor

Summary. The manuscript presents a compositional neural semantic parser for mapping sentences to graph-based meaning representations. It claims to achieve, for the first time, competitive accuracies across a diverse range of graphbanks (DM, PAS, PSD, AMR 2015, EDS) without per-graphbank hand-engineering; adding BERT embeddings and multi-task learning further improves results and sets new state-of-the-art numbers on those five graphbanks.

Significance. If the central claim holds and the compositional inductive bias (rather than multi-task learning or BERT alone) is shown to drive cross-graphbank generalization, the work would meaningfully reduce the need for graphbank-specific engineering in semantic parsing. The multi-task regime across graphbanks is a clear strength if properly isolated.

major comments (3)

[Abstract, §4] Abstract and §4 (experimental setup): the claim that the parser is 'compositional' and that this property enables competitive cross-graphbank performance is not isolated from the multi-task learning regime. No ablation is reported that removes the compositional components while retaining the shared encoder/decoder and joint training; without this, the 'for the first time' result could be explained by multi-task learning alone.
[§3] §3 (model architecture): the description of the compositional mechanism (e.g., how the decoder enforces graph compositionality) must be shown to differ substantively from a standard encoder-decoder that simply outputs graphs; if the compositionality is only in the output construction and not an enforced inductive bias, the generalization claim is at risk.
[Table 2, §5] Table 2 / §5 (results): the reported gains from multi-task learning versus single-graphbank training are not broken down per graphbank; without per-graphbank ablations it is impossible to verify that the model is not implicitly tuned to the collection of graphbanks used.

minor comments (2)

[§2] §2 (related work): a brief comparison table of prior single-graphbank parsers versus the proposed multi-graphbank approach would improve readability.
[§3] Notation in §3: the distinction between the compositional decoder state and the standard LSTM/Transformer state should be made explicit with an equation or diagram.

Simulated Author's Rebuttal

3 responses · 0 unresolved

Thank you for the constructive feedback. We address each major comment below, outlining planned revisions to strengthen the isolation of our claims and the clarity of the experimental results.

read point-by-point responses

Referee: [Abstract, §4] Abstract and §4 (experimental setup): the claim that the parser is 'compositional' and that this property enables competitive cross-graphbank performance is not isolated from the multi-task learning regime. No ablation is reported that removes the compositional components while retaining the shared encoder/decoder and joint training; without this, the 'for the first time' result could be explained by multi-task learning alone.

Authors: We agree that an ablation isolating the compositional inductive bias from the multi-task regime would strengthen the central claim. Our architecture uses a shared encoder/decoder with compositional decoding rules, but to directly address this we will add a new ablation in the revised version: a non-compositional decoder variant trained under the identical multi-task setup. This will help quantify the contribution of compositionality to cross-graphbank generalization. revision: yes
Referee: [§3] §3 (model architecture): the description of the compositional mechanism (e.g., how the decoder enforces graph compositionality) must be shown to differ substantively from a standard encoder-decoder that simply outputs graphs; if the compositionality is only in the output construction and not an enforced inductive bias, the generalization claim is at risk.

Authors: Section 3 describes the decoder as applying graph-specific composition rules (e.g., node/edge construction constraints) that enforce structural inductive biases during generation, unlike standard encoder-decoders that generate sequences without such constraints. We will revise §3 to explicitly contrast these mechanisms with standard seq2seq output construction and add examples illustrating the enforced bias. revision: yes
Referee: [Table 2, §5] Table 2 / §5 (results): the reported gains from multi-task learning versus single-graphbank training are not broken down per graphbank; without per-graphbank ablations it is impossible to verify that the model is not implicitly tuned to the collection of graphbanks used.

Authors: We will expand Table 2 and the corresponding discussion in §5 to report per-graphbank performance for single-graphbank training versus the multi-task setting. This breakdown will allow direct verification of gains on each individual graphbank. revision: yes

Circularity Check

0 steps flagged

No circularity in derivation chain

full rationale

The paper presents an empirical claim that a compositional neural semantic parser, enhanced with BERT embeddings and multi-task learning, achieves competitive accuracies and new state-of-the-art results on DM, PAS, PSD, AMR 2015 and EDS. No derivation chain, equations, or predictions are exhibited in the provided text that reduce to inputs by construction. No self-definitional steps, fitted parameters renamed as predictions, or load-bearing self-citations appear. The result is externally falsifiable via held-out graphbank evaluation and is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No free parameters, axioms, or invented entities can be identified from the abstract alone.

pith-pipeline@v0.9.0 · 5573 in / 913 out tokens · 21576 ms · 2026-05-25T14:53:51.233946+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

34 extracted references · 34 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...

work page
[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...

work page
[3]

Lasha Abzianidze, Johannes Bjerva, Kilian Evang, Hessel Haagsma, Rik van Noord, Pierre Ludmann, Duc-Duy Nguyen, and Johan Bos. 2017. http://aclweb.org/anthology/E17-2039 The parallel meaning bank: Towards a multilingual corpus of translations annotated with compositional meaning representations . In Proceedings of the 15th Conference of the European Chapt...

work page 2017
[4]

Z eljko Agi\' c , Alexander Koller, and Stephan Oepen. 2015. Semantic dependency graph parsing using tree approximations. In Proceedings of the 14th International Conference on Computational Semantics (IWCS)

work page 2015
[5]

Laura Banarescu, Claire Bonial, Shu Cai, Madalina Georgescu, Kira Griffitt, Ulf Hermjakob, Kevin Knight, Philipp Koehn, Martha Palmer, and Nathan Schneider. 2013. http://aclweb.org/anthology/W13-2322 A bstract M eaning R epresentation for Sembanking . In Proceedings of the 7th Linguistic Annotation Workshop and Interoperability with Discourse

work page 2013
[6]

Jan Buys and Phil Blunsom. 2017. https://doi.org/10.18653/v1/P17-1112 Robust incremental neural semantic graph parsing . In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics

work page doi:10.18653/v1/p17-1112 2017
[7]

Shu Cai and Kevin Knight. 2013. Smatch: an evaluation metric for semantic feature structures. In Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics

work page 2013
[8]

Yufei Chen, Weiwei Sun, and Xiaojun Wan. 2018. https://www.aclweb.org/anthology/P18-1038 Accurate SHRG -based semantic parsing . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 408--418, Melbourne, Australia. Association for Computational Linguistics

work page 2018
[9]

Ann Copestake, Dan Flickinger, Carl Pollard, and Ivan A Sag. 2005. Minimal recursion semantics: An introduction. Research on language and computation, 3(2-3):281--332

work page 2005
[10]

Hal Daum\'e III . 2007. http://aclweb.org/anthology/P07-1033 Frustratingly easy domain adaptation . In Proceedings of the 45th Annual Meeting of the Association of Computational Linguistics

work page 2007
[11]

Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. https://www.aclweb.org/anthology/N19-1423 BERT : Pre-training of deep bidirectional transformers for language understanding . In Proceedings of the 2019 Conference of the North A merican Chapter of the Association for Computational Linguistics: Human Language Technologies

work page 2019
[12]

Timothy Dozat and Christopher D. Manning. 2017. Deep biaffine attention for neural dependency parsing. In ICLR

work page 2017
[13]

Timothy Dozat and Christopher D. Manning. 2018. http://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

work page 2018
[14]

Rebecca Dridan and Stephan Oepen. 2011. Parser evaluation using elementary dependency matching. In Proceedings of the 12th International Conference on Parsing Technologies, pages 225--230

work page 2011
[15]

Dan Flickinger, Jan Haji c , Angelina Ivanova, Marco Kuhlmann, Yusuke Miyao, Stephan Oepen, and Daniel Zeman. 2017. http://hdl.handle.net/11234/1-1956 Open SDP 1.2 . LINDAT / CLARIN digital library at the Institute of Formal and Applied Linguistics ( \'U FAL ), Faculty of Mathematics and Physics, Charles University

work page 2017
[16]

Zettlemoyer

Matthew Gardner, Joel Grus, Mark Neumann, Oyvind Tafjord, Pradeep Dasigi, Nelson H S Liu, Matthew Peters, Michael Schmitz, and Luke S. Zettlemoyer. 2017. A deep semantic natural language processing platform

work page 2017
[17]

Jonas Groschwitz. 2019. http://www.coli.uni-saarland.de/ jonasg/thesis.pdf Methods for taking semantic graphs apart and putting them back together again . Ph.D. thesis, Macquarie University and Saarland University

work page 2019
[18]

Jonas Groschwitz, Meaghan Fowlie, Mark Johnson, and Alexander Koller. 2017. http://aclweb.org/anthology/W17-6810 A constrained graph algebra for semantic parsing with AMRs . In Proceedings of the 12th International Conference on Computational Semantics (IWCS)

work page 2017
[19]

Jonas Groschwitz, Matthias Lindemann, Meaghan Fowlie, Mark Johnson, and Alexander Koller. 2018. http://aclweb.org/anthology/P18-1170 A MR Dependency Parsing with a Typed Semantic Algebra . In Proceedings of ACL

work page 2018
[20]

Daniel Hershcovich, Omri Abend, and Ari Rappoport. 2018. http://aclweb.org/anthology/P18-1035 Multitask parsing across semantic representations . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics

work page 2018
[21]

Eliyahu Kiperwasser and Yoav Goldberg. 2016. http://aclweb.org/anthology/Q16-1023 S imple and Accurate Dependency Parsing Using Bidirectional LSTM Feature Representations . Transactions of the Association for Computational Linguistics, 4:313--327

work page 2016
[22]

Marco Kuhlmann and Stephan Oepen. 2016. http://aclweb.org/anthology/J16-4009 Towards a catalogue of linguistic graph banks . Computational Linguistics, 42(4):819--827

work page 2016
[23]

Wei Lu, Hai Leong Chieu, and Jonathan L \"o fgren. 2016. A general regularization framework for domain adaptation. In Proceedings of EMNLP

work page 2016
[24]

Chunchuan Lyu and Ivan Titov. 2018. http://aclweb.org/anthology/P18-1037 A MR Parsing as Graph Prediction with Latent Alignment . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics

work page 2018
[25]

Manning, Mihai Surdeanu, John Bauer, Jenny Finkel, Steven J

Christopher D. Manning, Mihai Surdeanu, John Bauer, Jenny Finkel, Steven J. Bethard, and David McClosky. 2014. The stanford corenlp natural language processing toolkit. In Proceedings of 52nd Annual Meeting of the Association for Computational Linguistics: System Demonstrations

work page 2014
[26]

Joakim Nivre, Mitchell Abrams, Z eljko Agi \'c , Lars Ahrenberg, Lene Antonsen, Katya Aplonova, Maria Jesus Aranzabe, et al. 2018. http://hdl.handle.net/11234/1-2895 Universal dependencies 2.3 . LINDAT / CLARIN digital library at the Institute of Formal and Applied Linguistics ( \'U FAL ), Faculty of Mathematics and Physics, Charles University

work page 2018
[27]

Stephan Oepen, Marco Kuhlmann, Yusuke Miyao, Daniel Zeman, Silvie Cinkov\' a , Dan Flickinger, Jan Haji c , and Zde n ka Ure s ov\' a . 2015. http://aclweb.org/anthology/S15-2153 Semeval 2015 task 18: Broad-coverage semantic dependency parsing . In Proceedings of the 9th International Workshop on Semantic Evaluation (SemEval 2015)

work page 2015
[28]

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 ACL

work page 2017
[29]

Hao Peng, Sam Thomson, Swabha Swayamdipta, and Noah A. Smith. 2018. https://doi.org/10.18653/v1/N18-1135 Learning joint semantic parsers from disjoint data . In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 1492--1502

work page doi:10.18653/v1/n18-1135 2018
[30]

Jeffrey Pennington, Richard Socher, and Christopher D. Manning. 2014. Glove: Global vectors for word representation. In Empirical Methods in Natural Language Processing (EMNLP)

work page 2014
[31]

Gabriel Stanovsky and Ido Dagan. 2018. http://aclweb.org/anthology/D18-1263 Semantics as a foreign language . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing

work page 2018
[32]

Sara Stymne, Miryam de Lhoneux, Aaron Smith, and Joakim Nivre. 2018. http://aclweb.org/anthology/P18-2098 Parser training with heterogeneous treebanks . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics

work page 2018
[33]

Tao Yu, Rui Zhang, Kai Yang, Michihiro Yasunaga, Dongxu Wang, Zifan Li, James Ma, Irene Li, Qingning Yao, Shanelle Roman, Zilin Zhang, and Dragomir Radev. 2018. http://aclweb.org/anthology/D18-1425 Spider: A large-scale human-labeled dataset for complex and cross-domain semantic parsing and text-to-sql task . In Proceedings of the 2018 Conference on Empir...

work page 2018
[34]

Sheng Zhang, Xutai Ma, Kevin Duh, and Benjamin Van Durme. 2019. AMR parsing as sequence-to-graph transduction. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (ACL)

work page 2019

[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...

work page

[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...

work page

[3] [3]

Lasha Abzianidze, Johannes Bjerva, Kilian Evang, Hessel Haagsma, Rik van Noord, Pierre Ludmann, Duc-Duy Nguyen, and Johan Bos. 2017. http://aclweb.org/anthology/E17-2039 The parallel meaning bank: Towards a multilingual corpus of translations annotated with compositional meaning representations . In Proceedings of the 15th Conference of the European Chapt...

work page 2017

[4] [4]

Z eljko Agi\' c , Alexander Koller, and Stephan Oepen. 2015. Semantic dependency graph parsing using tree approximations. In Proceedings of the 14th International Conference on Computational Semantics (IWCS)

work page 2015

[5] [5]

Laura Banarescu, Claire Bonial, Shu Cai, Madalina Georgescu, Kira Griffitt, Ulf Hermjakob, Kevin Knight, Philipp Koehn, Martha Palmer, and Nathan Schneider. 2013. http://aclweb.org/anthology/W13-2322 A bstract M eaning R epresentation for Sembanking . In Proceedings of the 7th Linguistic Annotation Workshop and Interoperability with Discourse

work page 2013

[6] [6]

Jan Buys and Phil Blunsom. 2017. https://doi.org/10.18653/v1/P17-1112 Robust incremental neural semantic graph parsing . In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics

work page doi:10.18653/v1/p17-1112 2017

[7] [7]

Shu Cai and Kevin Knight. 2013. Smatch: an evaluation metric for semantic feature structures. In Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics

work page 2013

[8] [8]

Yufei Chen, Weiwei Sun, and Xiaojun Wan. 2018. https://www.aclweb.org/anthology/P18-1038 Accurate SHRG -based semantic parsing . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 408--418, Melbourne, Australia. Association for Computational Linguistics

work page 2018

[9] [9]

Ann Copestake, Dan Flickinger, Carl Pollard, and Ivan A Sag. 2005. Minimal recursion semantics: An introduction. Research on language and computation, 3(2-3):281--332

work page 2005

[10] [10]

Hal Daum\'e III . 2007. http://aclweb.org/anthology/P07-1033 Frustratingly easy domain adaptation . In Proceedings of the 45th Annual Meeting of the Association of Computational Linguistics

work page 2007

[11] [11]

Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. https://www.aclweb.org/anthology/N19-1423 BERT : Pre-training of deep bidirectional transformers for language understanding . In Proceedings of the 2019 Conference of the North A merican Chapter of the Association for Computational Linguistics: Human Language Technologies

work page 2019

[12] [12]

Timothy Dozat and Christopher D. Manning. 2017. Deep biaffine attention for neural dependency parsing. In ICLR

work page 2017

[13] [13]

Timothy Dozat and Christopher D. Manning. 2018. http://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

work page 2018

[14] [14]

Rebecca Dridan and Stephan Oepen. 2011. Parser evaluation using elementary dependency matching. In Proceedings of the 12th International Conference on Parsing Technologies, pages 225--230

work page 2011

[15] [15]

Dan Flickinger, Jan Haji c , Angelina Ivanova, Marco Kuhlmann, Yusuke Miyao, Stephan Oepen, and Daniel Zeman. 2017. http://hdl.handle.net/11234/1-1956 Open SDP 1.2 . LINDAT / CLARIN digital library at the Institute of Formal and Applied Linguistics ( \'U FAL ), Faculty of Mathematics and Physics, Charles University

work page 2017

[16] [16]

Zettlemoyer

Matthew Gardner, Joel Grus, Mark Neumann, Oyvind Tafjord, Pradeep Dasigi, Nelson H S Liu, Matthew Peters, Michael Schmitz, and Luke S. Zettlemoyer. 2017. A deep semantic natural language processing platform

work page 2017

[17] [17]

Jonas Groschwitz. 2019. http://www.coli.uni-saarland.de/ jonasg/thesis.pdf Methods for taking semantic graphs apart and putting them back together again . Ph.D. thesis, Macquarie University and Saarland University

work page 2019

[18] [18]

Jonas Groschwitz, Meaghan Fowlie, Mark Johnson, and Alexander Koller. 2017. http://aclweb.org/anthology/W17-6810 A constrained graph algebra for semantic parsing with AMRs . In Proceedings of the 12th International Conference on Computational Semantics (IWCS)

work page 2017

[19] [19]

Jonas Groschwitz, Matthias Lindemann, Meaghan Fowlie, Mark Johnson, and Alexander Koller. 2018. http://aclweb.org/anthology/P18-1170 A MR Dependency Parsing with a Typed Semantic Algebra . In Proceedings of ACL

work page 2018

[20] [20]

Daniel Hershcovich, Omri Abend, and Ari Rappoport. 2018. http://aclweb.org/anthology/P18-1035 Multitask parsing across semantic representations . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics

work page 2018

[21] [21]

Eliyahu Kiperwasser and Yoav Goldberg. 2016. http://aclweb.org/anthology/Q16-1023 S imple and Accurate Dependency Parsing Using Bidirectional LSTM Feature Representations . Transactions of the Association for Computational Linguistics, 4:313--327

work page 2016

[22] [22]

Marco Kuhlmann and Stephan Oepen. 2016. http://aclweb.org/anthology/J16-4009 Towards a catalogue of linguistic graph banks . Computational Linguistics, 42(4):819--827

work page 2016

[23] [23]

Wei Lu, Hai Leong Chieu, and Jonathan L \"o fgren. 2016. A general regularization framework for domain adaptation. In Proceedings of EMNLP

work page 2016

[24] [24]

Chunchuan Lyu and Ivan Titov. 2018. http://aclweb.org/anthology/P18-1037 A MR Parsing as Graph Prediction with Latent Alignment . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics

work page 2018

[25] [25]

Manning, Mihai Surdeanu, John Bauer, Jenny Finkel, Steven J

Christopher D. Manning, Mihai Surdeanu, John Bauer, Jenny Finkel, Steven J. Bethard, and David McClosky. 2014. The stanford corenlp natural language processing toolkit. In Proceedings of 52nd Annual Meeting of the Association for Computational Linguistics: System Demonstrations

work page 2014

[26] [26]

Joakim Nivre, Mitchell Abrams, Z eljko Agi \'c , Lars Ahrenberg, Lene Antonsen, Katya Aplonova, Maria Jesus Aranzabe, et al. 2018. http://hdl.handle.net/11234/1-2895 Universal dependencies 2.3 . LINDAT / CLARIN digital library at the Institute of Formal and Applied Linguistics ( \'U FAL ), Faculty of Mathematics and Physics, Charles University

work page 2018

[27] [27]

Stephan Oepen, Marco Kuhlmann, Yusuke Miyao, Daniel Zeman, Silvie Cinkov\' a , Dan Flickinger, Jan Haji c , and Zde n ka Ure s ov\' a . 2015. http://aclweb.org/anthology/S15-2153 Semeval 2015 task 18: Broad-coverage semantic dependency parsing . In Proceedings of the 9th International Workshop on Semantic Evaluation (SemEval 2015)

work page 2015

[28] [28]

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 ACL

work page 2017

[29] [29]

Hao Peng, Sam Thomson, Swabha Swayamdipta, and Noah A. Smith. 2018. https://doi.org/10.18653/v1/N18-1135 Learning joint semantic parsers from disjoint data . In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 1492--1502

work page doi:10.18653/v1/n18-1135 2018

[30] [30]

Jeffrey Pennington, Richard Socher, and Christopher D. Manning. 2014. Glove: Global vectors for word representation. In Empirical Methods in Natural Language Processing (EMNLP)

work page 2014

[31] [31]

Gabriel Stanovsky and Ido Dagan. 2018. http://aclweb.org/anthology/D18-1263 Semantics as a foreign language . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing

work page 2018

[32] [32]

Sara Stymne, Miryam de Lhoneux, Aaron Smith, and Joakim Nivre. 2018. http://aclweb.org/anthology/P18-2098 Parser training with heterogeneous treebanks . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics

work page 2018

[33] [33]

Tao Yu, Rui Zhang, Kai Yang, Michihiro Yasunaga, Dongxu Wang, Zifan Li, James Ma, Irene Li, Qingning Yao, Shanelle Roman, Zilin Zhang, and Dragomir Radev. 2018. http://aclweb.org/anthology/D18-1425 Spider: A large-scale human-labeled dataset for complex and cross-domain semantic parsing and text-to-sql task . In Proceedings of the 2018 Conference on Empir...

work page 2018

[34] [34]

Sheng Zhang, Xutai Ma, Kevin Duh, and Benjamin Van Durme. 2019. AMR parsing as sequence-to-graph transduction. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (ACL)

work page 2019