Universal Dependencies v2: An Evergrowing Multilingual Treebank Collection

By Joakim Nivre, Marie-Catherine de Marneffe, Filip Ginter, Jan Hajic, Jan Hajivc, Christopher D. Manning, Sampo Pyysalo, Sebastian Schuster, Francis M. Tyers, Daniel Zeman. International Conference on Language Resources and Evaluation, 2020

657 citations recorded by the source corpus.

Universal Dependencies is an open community effort to create cross-linguistically consistent treebank annotation for many languages within a dependency-based lexicalist framework. The annotation consists in a linguistically motivated word segmentation; a morphological layer comprising lemmas, universal part-of-speech tags, and standardized morphological features; and a syntactic layer focusing on syntactic relations between predicates, arguments and modifiers. In this paper, we describe version 2 of the universal guidelines (UD v2), discuss the major changes from UD v1 to UD v2, and give an overview of the currently available treebanks for 90 languages.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/b805693c17961af2cc7f859c1a54320b26036f46

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Five sources of bias in natural language processing (2021, background): “The shortage of syntactic resources has since been addressed by the Universal Dependency Project (Nivre et al., 2020).”
Global Explainability of BERT-Based Evaluation Metrics by Disentangling along Linguistic Factors (2021, methodology): “, 2017) and increase morphological information in the original embeddings: (a) first, we produce two morphological lexicons based on words from WMT and STS, each containing word pairs with identical UD morphological tags (Nivre et al., 2020) (see Table 2).”
Challenges and Strategies in Cross-Cultural NLP (2022, background): “Examples for such communities include Universal Dependencies, dedicated to manual annotation of morphosyntactic datasets for over one hundred languages, covering the highest typological diversity to date among such datasets (Nivre et al., 2020).”
Substructure Substitution: Structured Data Augmentation for NLP (2021, methodology): “We conduct our experiments using the Universal Dependencies (UD; Nivre et al., 2016, 2020) 4 dataset.”
Can Language Models Encode Perceptual Structure Without Grounding? A Case Study in Color (2021, background): “…the other hand, indicates that the degree of speciﬁcation offered by the POS tagset might be too coarse to meaningfully differentiate be-tween color terms, e.g. nouns can occur in a variety of DRELs such as subjects, objects, oblique modiﬁers (per the Universal Depende-cies (Nivre et al., 2020)).”
Conditional probing: measuring usable information beyond a baseline (2021, background): “We predict ﬁve linguistic properties Y : (i) upos : coarse-grained (17-tag) part-of-speech tags (Nivre et al., 2020), (ii) xpos : ﬁne-grained English-speciﬁc part-of-speech tags, (iii) dep rel : the label on the Universal Dependencies edge that governs the word, (iv) ner : named entities, and (v)…”
A Survey on Recent Approaches for Natural Language Processing in Low-Resource Scenarios (2020, background): “(2020) study languages that have less than 10k labeled tokens in the Universal Dependency project (Nivre et al., 2020) and Loubser and Puttkammer (2020) report that most available datasets for South”
Is Supervised Syntactic Parsing Beneficial for Language Understanding Tasks? An Empirical Investigation (2020, background): “…based on a variety of syntactic formalisms (Charniak, 1996; Taylor et al., 2003; De Marneffe et al., 2006; Hock-enmaier and Steedman, 2007; Nivre et al., 2016, 2020, inter alia ), has been the beating heart of NLP pipelines for decades (Klein and Manning, 2003; Chen and Manning, 2014;…”
Finding Universal Grammatical Relations in Multilingual BERT (2020, methodology): “• Representing a dependency by the difference of the head and dependent vectors in the syntactic space, we show that mBERT represents dependency clusters that largely overlap with the dependency taxonomy of Universal Dependencies (UD) (Nivre et al., 2020); see Figure 1.”
Evaluating the Morphosyntactic Well-formedness of Generated Texts (2021, methodology): “Both rele- 4Different syntactic formalisms could be applicable, but we work with the (modified) Universal Dependencies formalism (Nivre et al., 2020) due to its simplicity, widespread familiarity and its use in a variety of multilingual resources.”
Square One Bias in NLP: Towards a Multi-Dimensional Exploration of the Research Manifold (2022, background): “The Universal Dependencies project (Nivre et al., 2020) is motivated by the observation that not all dependency formalisms are easily”
How Good is Your Tokenizer? On the Monolingual Performance of Multilingual Language Models (2020, methodology): “We do not specifically preprocess any of the UD (Nivre et al., 2016, 2020; Zeman et al., 2020) treebanks.”
Evaluating the Quality of a Corpus Annotation Scheme Using Pretrained Language Models (2024, background): “Universal Dependencies (UD) (De Marneffe et al., 2021) project is a framework that provides tree-banks in a dependency grammar format (Bauer, 1979; Debusmann, 2000; De Marneffe and Nivre, 2019) in more than 100 languages (Nivre et al., 2020).”
A Learning-Based Dependency to Constituency Conversion Algorithm for the Turkish Language (2022, background): “With the Universal Depen-dencies’ (Nivre et al., 2020) attempt to offer a framework and set of rules that can be applied to a wide range of different languages, an impressively comprehensive multilingual dependency treebank 1 was created to offer much needed data -especially for low resource languag”
Construction and educational application of a linguistically grounded dependency treebank for Uyghur (2025, unknown): “Currently, the Universal Dependencies (UD) framework serves as the de facto standard for syntactic annotation due to its cross-linguistic consistency (de Marneffe et al., 2021; Nivre et al., 2020).”
Syntactic dependency length shaped by strategic memory allocation (2024, methodology): “We examine our hypothesis using the corpora taken from Universal Dependencies (UD) release 2.11 (Nivre et al., 2020), as described in Table 1.”