Universal Stanford dependencies: A cross-linguistic typology

By Marie-Catherine de Marneffe, Timothy Dozat, Natalia Silveira, Katri Haverinen, Filip Ginter, Joakim Nivre, Christopher D. Manning. International Conference on Language Resources and Evaluation, 2014

556 citations recorded by the source corpus.

Revisiting the now de facto standard Stanford dependency representation, we propose an improved taxonomy to capture grammatical relations across languages, including morphologically rich ones. We suggest a two-layered taxonomy: a set of broadly attested universal grammatical relations, to which language-specific relations can be added. We emphasize the lexicalist stance of the Stanford Dependencies, which leads to a particular, partially new treatment of compounding, prepositions, and morphology. We show how existing dependency schemes for several languages map onto the universal taxonomy proposed here and close with consideration of practical implications of dependency representation choices for NLP applications, in particular parsing.

Subfields: Artificial Intelligence

View Paper: https://www.semanticscholar.org/paper/5b117fb14e3622d8c675c1c97b8b29e175dea194

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

Large-scale evidence of dependency length minimization in 37 languages (2015, methodology): “Thirty-six of the corpora follow widely recognized standards for dependency analysis (25, 26); the remaining corpus (Mandarin Chinese) uses its own system that is nonetheless similar to the standards [see Table S1 for details on each corpus].”
Posterior calibration and exploratory analysis for natural language processing models (2015, background): “…subset of the Annotated New York Times LDC corpus (Sand-haus, 2008), which includes articles about world 10 Syntactic relations are Universal Dependencies (de Marneffe et al., 2014); more details for the extraction rules are in the appendix. news from the years 1987 to 2007 (details in appendix).”
Combining Natural Logic and Shallow Reasoning for Question Answering (2016, methodology): “We adapt this mapping to yield the relation induced by inserting a given dependency edge, corresponding to our deletions in search; we also convert the mapping to use Universal Dependencies (de Marneffe et al., 2014).”
Demographic Dialectal Variation in Social Media: A Case Study of African-American English (2016, unknown): “We also considered follow-up works de Marneffe et al. (2014) and http://universaldependencies.org/, deferring to de Marneffe and Manning (2008) when in conﬂict.”
Morphological Segmentation and OPUS for Finnish-English Machine Translation (2015, methodology): “This pipeline integrates all pre-processing steps that are necessary for data-driven dependency parsing including tokenization, morphological analyses and part-ofspeech tagging, and produces dependency analyses in a minor variant of the Stanford Dependencies scheme (de Marneffe et al., 2014).”
Mimicking Word Embeddings using Subword RNNs (2017, methodology): “phosyntactic annotations through the Universal Dependencies (UD) corpus (De Marneffe et al., 2014). For example, Faruqui et al. (2016) propose a graph-based technique for propagating typelevel morphological information across a lexicon, improving token-level morphosyntactic tagging in 11 languages, ”
SMT and Hybrid systems of the QTLeap project in the WMT16 IT-task (2016, methodology): “After running the shallow processing tools, the dependency output of the parser is converted into Universal Dependencies (UD) (de Marneffe et al., 2014).”
Generating Semantically Precise Scene Graphs from Textual Descriptions for Improved Image Retrieval (2015, background): “While one of the principles of the Universal Dependencies representation is the primacy of content words (de Marneffe et al., 2014), light nouns are treated like any other noun.”
Interpretable Visual Question Answering by Reasoning on Dependency Trees (2018, background): “One of the most widely used head-dependent relation sets is Universal Dependencies (UD) [38].”
SPICE: Semantic Propositional Image Caption Evaluation (2016, methodology): “An example of the resulting dependency syntax tree, using Universal Dependency relations [20], is shown in Figure 1 top.”
Deep Multitask Learning for Semantic Dependency Parsing (2017, background): “Even syntactic formalisms are moving toward graphs (de Marneffe et al., 2014).”
Visual Question Reasoning on General Dependency Tree (2018, methodology): “One of the most widely used head-dependent relation sets is the Universal Dependencies(UD) [7].”
The GUM corpus: creating multilayer resources in the classroom (2016, methodology): “Currently a conversion adding Universal Stanford Dependencies (de Marneffe et al. 2014) to the Typed Dependencies is planned, as is the addition of a sentence mood/rough speech act layer.”
A Gold Standard Dependency Corpus for English (2014, background): “The SD standard, which is English-centric, is also under revision for cross-linguistic adequacy, as described in (de Marneffe et al., 2014).”
Universal Dependencies v2: An Evergrowing Multilingual Treebank Collection (2020, methodology): “The annotation scheme is based on an evolution of (universal) Stanford dependencies (de Marneffe et al., 2006; de Marneffe and Manning, 2008; de Marneffe et al., 2014), Google universal part-of-speech tags (Petrov et al., 2012), and the Interset interlingua for morphosyn-tactic tagsets (Zeman,…”
NegBio: a high-performance tool for negation and uncertainty detection in radiology reports (2017, methodology): “The universal dependencies were then obtained by applying the Stanford dependencies converter on the parse tree with the CCProcessed and Universal option 21,29 .”
Deep learning for extracting protein-protein interactions from biomedical literature (2017, methodology): “The universal dependencies were then obtained by applying the Stanford dependencies converter on the parse tree with the CCPro-cessed and Universal options (De Marneffe et al., 2014).”