Mova · papers by stage
tokens → tree
Building the Universal Dependencies tree: the transition parser, the treebanks, the grammars of English as rule sources. The 50 works of Mova's own list that this stage uses · page 2 of 2; measured marks a work whose effect Mova measured in its own experiments (3 here). Each title opens our card: the summary, how Mova uses the work, its references and who cites it. What the stage does is on Mova's pipeline page; every stage is on the main list.
- The BEA-2019 Shared Task on Grammatical Error CorrectionChristopher Bryant et al. · 2019
What Mova took: en/seeds/errors/errant-taxonomy.md: the table of the most frequent edit types in W&I+LOCNESS, FCE and NUCLE comes from this paper, and is used to prioritise which error types get deterministic rules; its finding that GEC systems…
- Estimating POS Annotation Consistency of Different Treebanks in a LanguageAkshay Aggarwal, Daniel Zeman · 2020
What Mova took: Summarised as one of five methods for detecting annotation errors automatically in the seed en/seeds/errors/consistency-detection-methods.md, which guides how the grammar expert system (en/src/expert.rs) and its rules are…
- Classifying Syntactic Errors in Learner LanguageLeshem Choshen, Dmitry Nikolaev, Yevgeni Berzak, Omri Abend · 2020
What Mova took: en/seeds/errors/errant-taxonomy.md: cited as the complement to ERRANT — SERCL classifies syntactically a large share of what ERRANT leaves as OTHER, and supports Mova's view that error types with a trace in the UD tree can be…
- Identifying and Handling Cross-Treebank Inconsistencies in UD: A Pilot StudyTillmann Dönicke, Xiang Yu, Jonas Kuhn · 2020
What Mova took:
en/seeds/errors/consistency-detection-methods.md: the arc-direction (MBD) method is listed as one of five ways to find annotation errors automatically. Mova adopted its logic as a family of direction rules:flatandconjgo… - Universal DependenciesMarie-Catherine de Marneffe et al. · 2021
What Mova took: en/seeds/errors/numbered-entities-nummod.md: cited in the history of conventions for numbers after a noun ("page 394", "Route 66"): the article recommended
nmod, while the current UD guideline usesflat; the rule card… - Coordinate Constructions in English Enhanced Universal Dependencies: Analysis and Computational ModelingStefan Grünewald, Prisca Piccirilli, Annemarie Friedrich · 2021
What Mova took:
en/seeds/errors/coordination-structure.md: cited in the seed on UD v2 coordination (the first conjunct is the head, andccattaches to the following conjunct) for the observation that the Enhanced UD converter fails… - Minor changes make a difference: a case study on the consistency of UD-based dependency parsersDmytro Kalpakchi, Johan Boye · 2021
What Mova took:
en/seeds/errors/consistency-detection-methods.md— listed as method 5 ("robustness to small changes") among automatic ways to find annotation errors: places where the parse flips under a trivial substitution are candidates for… - Mischievous nominal constructions in Universal DependenciesNathan Schneider, Amir Zeldes · 2021
What Mova took:
en/seeds/errors/titles-and-company-suffixes.md: ruleen.errors.title-descfor the English grammar expert system (en/src/expert.rs) — titles before names and company suffixes (Inc., Corp., Ltd.) should benmod:desc; the… - A Dependency Treebank of Spoken Second Language EnglishKristopher Kyle et al. · 2022
What Mova took:
en/seeds/errors/literal-annotation-learner.md— cited as evidence that learner language needs in-domain data (a little L2 data lifts LAS from 0.876 to 0.935), in the note establishing that Mova annotates erroneous text… - Conjunct Lengths in English, Dependency Length Minimization, and Dependency Structure of CoordinationAdam Przepiórkowski, Michał Woźniak · 2023
What Mova took:
en/seeds/errors/coordination-structure.md— cited in the seed on UD coordination as a caveat: UD's structure is asymmetric (the first conjunct is the head), and this paper shows such asymmetric models fail to explain… - Subject-verb agreement with Seq2Seq transformers: Bigger is better, but still not bestMichael Wilson, Zhenghao Zhou, Robert Frank · 2023
What Mova took:
en/seeds/errors/plural-subject-singular-verb.md(ruleen.errors.plural-subject-vbz) anden/seeds/errors/singular-subject-plural-verb.md(ruleen.errors.singular-subject-vbp): agreement rules of the English grammar expert… - Assessing the influence of attractor-verb distance on grammatical agreement in humans and language modelsChristos Zacharopoulos, Théo Desbordes, Mathias Sablé-Meyer · 2023
What Mova took:
en/seeds/errors/attraction-of-phrase.md: ruleen.errors.attraction-ofin the English grammar expert system (en/src/expert.rs) — a singular head noun with a plural of-phrase followed by a plural verb… - Are UD Treebanks Getting More Consistent? A Report Card for English UDAmir Zeldes, Nathan Schneider · 2023
What Mova took:
english-treebank-conventions.md: divergences between EWT and GUM are treated as conventions, not rule errors; a much higher violation count on GUM than on EWT signals a convention difference. - «Are you Afraid of Ghosts?» A Proposal for Busting Predicate Ellipsis in Universal DependenciesClaudia Corbetta, Federica Iurescia, Marco Carlo Passarotti · 2025
What Mova took: en/seeds/errors/gapping-orphan.md: rule
en.errors.orphan-head— in gapping ("Marie went to Paris and Miriam to Prague") anorphanmust hang on the promoted remnant (a conjunct, root,advclorparataxis); otherwise the… - Syntax of referents of relative markers: Evidence from a corpus of learner EnglishIzabela Czerniak, Debopam Das · 2025
What Mova took: en/seeds/errors/relative-pronoun-in-clause.md: cited as background on antecedent functions for rule
en.errors.relpron-inside-clause— a relative pronoun must be a member of the relative clause (nsubj,obj,obl…) under its… - ICLE-RC: International Corpus of Learner English for Relative ClausesDebopam Das, Izabela Czerniak, Peter Bourgonje · 2025
What Mova took: en/seeds/errors/relative-pronoun-in-clause.md: cited for the inventory of relative markers (that, wh-forms, zero) and for the observation that learners also use that in non-restrictive relatives; supports rule…
- Negation in Universal DependenciesJamie Yates Findlay, Dag Trygve Truslew Haug · 2025
What Mova took:
en/seeds/errors/negation-not.md: grounds the error ruleen.errors.not-part(not/n't must be PART, XPOS RB, Polarity=Neg; the oldnegrelation no longer exists). The paper's observation that EWT lacked Polarity=Neg in UD… - Assessing the Agreement Competence of Large Language ModelsAlba Táboas García, Leo Wanner · 2025
What Mova took:
en/seeds/errors/attraction-of-phrase.md: cited as evidence that attractor constructions are the hardest for LLMs. This motivates the warning rulesen.errors.attraction-ofanden.errors.attraction-of-auxcop, which flag a… - LLM Dependency Parsing with In-Context RulesMichael Ginn, Alexis Palmer · 2025
What Mova took:
en/seeds/errors/llm-annotator-failure-modes.md: one of the sources for the catalogue of typical errors LLM annotators make in UD. The paper backs item 6: LLM-generated rules tend to be too narrow… - Parsing the Switch: LLM-Based UD Annotation for Complex Code-Switched and Low-Resource LanguagesOlga Kellert et al. · 2025
What Mova took:
en/seeds/errors/dep-last-resort.md— cited as a source of the ruleen.errors.dep-last-resort: LLM annotators fall back todepon repetitions, ellipsis and contractions (the paper's prompt even alloweddepfor ellipsis)… - Targeted Syntactic Evaluation for Grammatical Error CorrectionAomi Koyama et al. · 2025
What Mova took:
en/seeds/errors/errant-taxonomy.md— cited alongside ERRANT as a finer-grained, CEFR-ordered view of error types (verb tense split into 15 subtypes). - Annotating Second Language in Universal Dependencies: a Review of Current Practices and Directions for Harmonized GuidelinesArianna Masciolini et al. · 2025
What Mova took:
en/seeds/errors/overregularized-verbs.md— source of the ruleen.errors.overregularized-typo: over-regularised forms (goed, buyed) are kept as written, with Typo=Yes, the lemma and tag of the intended word and the correct… - Step-by-step Instructions and a Simple Tabular Output Format Improve the Dependency Parsing Accuracy of LLMsHiroshi Matsuda, Chunpeng Ma, Masayuki Asahara · 2025
What Mova took:
en/seeds/errors/llm-annotator-failure-modes.md— cited as evidence in the catalogue of typical LLM annotation failures: formal errors (cycles are rare) and what helps (the UPOS → head → relation step order and tabular output). - Probing the Dynamics of Syntactic Ability Acquisition Throughout LLM PretrainingHiroshi Matsuda, Masayuki Asahara · 2026
What Mova took:
en/seeds/errors/clausal-subject-csubj.md— the finding that csubj is late and unstable in LLMs supports a rule flaggingnsubjon a VERB (a clausal subject should becsubj). - Encoding Logical Relations of Chinese Complex Sentences within the Universal Dependencies FrameworkHongpu Zhu, Hongzhi Xu · 2026
What Mova took:
en/seeds/errors/llm-annotator-failure-modes.md— a seed (a short declarative rule note) listing typical failure modes of LLM-based UD annotators, with a check for each. This paper is one of the sources for the failure mode…