Universal Dependencies Research Articles

Health sciences training for disability inclusion: the need to engage with emotion

Material aspects of disability inequality, such as access to quality services, housing and employment, are an urgent, worldwide concern. Less recognised though, are psychological layers of prejudice and discrimination, which...

Exploring language relations through syntactic distances and geographic proximity

Languages are grouped into families that share common linguistic traits. While this approach has been successful in understanding genetic relations between diverse languages, more analyses are needed to accurately quantify their relatedness, especially in less studied linguistic levels such as syntax. Here, we explore linguistic distances using series of parts of speech (POS) extracted from the Universal Dependencies dataset. Within an information-theoretic framework, we show that employing POS trigrams maximizes the possibility of capturing syntactic variations while being at the same time compatible with the amount of available data. Linguistic connections are then established by assessing pairwise distances based on the POS distributions. Intriguingly, our analysis reveals definite clusters that correspond to well known language families and groups, with exceptions explained by distinct morphological typologies. Furthermore, we obtain a significant correlation between language similarity and geographic distance, which underscores the influence of spatial proximity on language kinships.

PoeTree: Poetry Treebanks in Czech, English, French, German, Hungarian, Italian, Portuguese, Russian, Slovenian and Spanish

Abstract This article presents a set of standardised corpora of poetry comprising over 330,000 poems in ten languages (Czech, English, French, German, Hungarian, Italian, Portuguese, Russian, Slovenian, and Spanish). Each corpus has been deduplicated, enriched with Universal Dependencies, provided with additional metadata, and converted into a unified json structure.

Cross-lingual dependency parsing for a language with a unique script

Abstract Syntactic parsing is one of the areas in Natural Language Processing. The development of large-scale multilingual language models has enabled cross-lingual parsing approaches, which allows us to develop parsers for languages that do not have treebanks available. However, these approaches rely on the assumption that languages share orthographic representations and lexical entries. In this article, we investigate methods for developing a dependency parser for Xibe, a low-resource language that is written in a unique script. We first investigate lexicalized monolingual dependency parsing experiments to examine the effectiveness of word, part-of-speech, and character embeddings as well as pre-trained language models. Results show that character embeddings can significantly improve performance, while pre-trained language models decrease performance since they do not recognize the Xibe script. We also train delexicalized monolingual models, which yield competitive results to the best lexicalized model. Since the monolingual models are trained on a very small training set, we also investigate lexicalized and delexicalized cross-lingual models. We use six closely related languages as source language, which cover a wide range of scripts. In this setting, the delexicalized models achieve higher performance than lexicalized models. A final experiment shows that we can increase performance of the cross-lingual model by combining source languages and selecting the most similar sentences to Xibe as training set. However, all cross-lingual parsing results are still considerably lower than the monolingual model. We attribute the low performance of cross-lingual methods to syntactic and annotation differences as well as to the impoverished input of Universal Dependency Part-of-Speech tags that the delexicalized model has access to.

Spoken Spanish PoS tagging: gold standard dataset

AbstractThe development of a benchmark for part-of-speech (PoS) tagging of spoken dialectal European Spanish is presented, which will serve as the foundation for a future treebank. The benchmark is constructed using transcriptions of the Corpus Oral y Sonoro del Español Rural (COSER;“Audible corpus of spoken rural Spanish”) and follows the Universal Dependencies project guidelines. We describe the methodology used to create a gold standard, which serves to evaluate different state-of-the-art PoS taggers (spaCy, Stanza NLP, and UDPipe), originally trained on written data and to fine-tune and evaluate a model for spoken Spanish. It is shown that the accuracy of these taggers drops from 0.98$$-$$ - 0.99 to 0.94$$-$$ - 0.95 when tested on spoken data. Of these three taggers, the spaCy’s trf (transformers) and Stanza NLP models performed the best. Finally, the spaCy trf model is fine-tuned using our gold standard, which resulted in an accuracy of 0.98 for coarse-grained tags (UPOS) and 0.97 for fine-grained tags (FEATS). Our benchmark will enable the development of more accurate PoS taggers for spoken Spanish and facilitate the construction of a treebank for European Spanish varieties.

An accurate transformer-based model for transition-based dependency parsing of free word order languages

Transformer models are the state-of-the-art in Natural Language Processing (NLP) and the core of the Large Language Models (LLMs). We propose a transformer-based model for transition-based dependency parsing of free word order languages. We have performed experiments on five treebanks from the Universal Dependencies (UD) dataset version 2.12. Our experiments show that a transformer model, trained with the dynamic word embeddings performs better than a multilayer perceptron trained on the state-of-the-art static word embeddings even if the dynamic word embeddings have a vocabulary size ten times smaller than the static word embeddings. The results show that the transformer trained on dynamic word embeddings achieves an unlabeled attachment score (UAS) of 84.17% for Urdu language which is ≈3.6% and ≈1.9% higher than the UAS scores of 80.56857% and 82.26859% achieved by the multilayer perceptron (MLP) using two static state-of-the-art word embeddings. The proposed approach is investigated for Arabic, Persian and Uyghur languages, in addition to Urdu, for UAS scores and the results suggest that the proposed solution outperform the MLP-based approaches.

Old Permic Universal Dependencies Treebank

Old Permic, also known as Old Komi, is an extinct variety of Komi that was spoken in the late Middle Ages in the lower Vychegda river basin in Northeastern European Russia, in an area that currently is not Komi-speaking. This language variety is attested in fragmentary records from the 14th to 17th century written both in the Old Permic alphabet and in Cyrillic. These records are of significant importance for research on the history of the Komi language. Here we introduce our attempt towards a new Universal Dependencies treebank that will eventually contain the existing corpus of Old Permic in a structured and CoNLL-U annotated format. This will be the first time this material is being made openly available in digital format, and our contribution describes the current state of the art and remaining challenges.

Let's Do It Orderly: A Proposal for a Better Taxonomy of Adverbs in Universal Dependencies, and Beyond

Let's Do It Orderly: A Proposal for a Better Taxonomy of Adverbs in Universal Dependencies, and Beyond

Care dependency in older stroke patients with comorbidities: a latent profile analysis.

To explore latent profiles of care dependency in older stroke patients with comorbidities and to analyze the factors influencing different latent profiles. A total of 312 older ischemic stroke patients with comorbidities were included in the analysis. Latent Profile Analysis (LPA) was used to classify the participants into potential subgroups with different types of care dependency. The influencing factors of the classification of care dependency subgroups were determined using multivariate Logistic regression analysis. The care dependency score of older ischemic stroke patients with comorbidities was (51.35 ± 13.19), and the patients could be classified into 3 profiles, namely Universal dependency (24.0%), Moderate activity-social-learning dependency (28.0%), and Mild activity-social-learning dependency (48.0%); caregiver, BI at admission, and functional impairments were independent factors influencing care dependency (P < 0.05). There are three latent profiles of care dependency in older ischemic stroke patients with comorbidities. According to the characteristics of various populations, medical staff are able to implement specific interventions to lower the level of dependency and further improve the quality of life of patients.

Cross-linguistically consistent semantic and syntactic annotation of child-directed speech

AbstractCorpora of child speech and child-directed speech (CDS) have enabled major contributions to the study of child language acquisition, yet semantic annotation for such corpora is still scarce and lacks a uniform standard. Semantic annotation of CDS is particularly important for understanding the nature of the input children receive and developing computational models of child language acquisition. For example, under the assumption that children are able to infer meaning representations for (at least some of) the utterances they hear, the acquisition task is to learn a grammar that can map novel adult utterances onto their corresponding meaning representations, in the face of noise and distraction by other contextually possible meanings. To study this problem and to develop computational models of it, we need corpora that provide both adult utterances and their meaning representations, ideally using annotation that is consistent across a range of languages in order to facilitate cross-linguistic comparative studies. This paper proposes a methodology for constructing such corpora of CDS paired with sentential logical forms, and uses this method to create two such corpora, in English and Hebrew. The approach enforces a cross-linguistically consistent representation, building on recent advances in dependency representation and semantic parsing. Specifically, the approach involves two steps. First, we annotate the corpora using the Universal Dependencies (UD) scheme for syntactic annotation, which has been developed to apply consistently to a wide variety of domains and typologically diverse languages. Next, we further annotate these data by applying an automatic method for transducing sentential logical forms (LFs) from UD structures. The UD and LF representations have complementary strengths: UD structures are language-neutral and support consistent and reliable annotation by multiple annotators, whereas LFs are neutral as to their syntactic derivation and transparently encode semantic relations. Using this approach, we provide syntactic and semantic annotation for two corpora from CHILDES: Brown’s Adam corpus (English; we annotate $$\approx$$ ≈ 80% of its child-directed utterances), all child-directed utterances from Berman’s Hagar corpus (Hebrew). We verify the quality of the UD annotation using an inter-annotator agreement study, and manually evaluate the transduced meaning representations. We then demonstrate the utility of the compiled corpora through (1) a longitudinal corpus study of the prevalence of different syntactic and semantic phenomena in the CDS, and (2) applying an existing computational model of language acquisition to the two corpora and briefly comparing the results across languages.

Anisotropic Galvanomagnetic Effects in Single Cubic Crystals: A Theory and Its Verification.

A theory of anisotropic galvanomagnetic effects in single cubic crystals and its experimental verifications are presented for the current in the (001) plane. In contrast to the general belief that galvanomagnetic effects in single crystals are highly sensitive to many internal and external effects and have no universal features, the theory predicts universal angular dependencies of longitudinal and transverse resistivity and various characteristics when magnetization rotates in the (001) plane, the plane perpendicular to the current, and the plane containing the current and [001] direction. The universal angular dependencies are verified by experiments on Fe_{30}Co_{70} single cubic crystal film. The findings provide new avenues for fundamental research and applications of galvanomagnetic effects, because single crystals offer advantages over polycrystalline materials for band structure and crystallographic orientation engineering.

Streamlining event extraction with a simplified annotation framework.

Event extraction, grounded in semantic relationships, can serve as a simplified relation extraction. In this study, we propose an efficient open-domain event annotation framework tailored for subsequent information extraction, with a specific focus on its applicability to low-resource languages. The proposed event annotation method, which is based on event semantic elements, demonstrates substantial time-efficiency gains over traditional Universal Dependencies (UD) tagging. We show how language-specific pretraining outperforms multilingual counterparts in entity and relation extraction tasks and emphasize the importance of task- and language-specific fine-tuning for optimal model performance. Furthermore, we demonstrate the improvement of model performance upon integrating UD information during pre-training, achieving the F1 score of 71.16 and 60.43% for entity and relation extraction respectively. In addition, we showcase the usage of our extracted event graph for improving node classification in a retail banking domain. This work provides valuable guidance on improving information extraction and outlines a methodology for developing training datasets, particularly for low-resource languages.

Morphosyntactic Annotation in Literary Stylometry

This article investigates the stylometric usefulness of morphosyntactic annotation. Focusing on the style of literary texts, it argues that including morphosyntactic annotation in analyses of style has at least two important advantages: (1) maintaining a topic agnostic approach and (2) providing input variables that are interpretable in traditional grammatical terms. This study demonstrates how widely available Universal Dependency parsers can generate useful morphological and syntactic data for texts in a range of languages. These data can serve as the basis for input features that are strongly informative about the style of individual novels, as indicated by accuracy in classification tests. The interpretability of such features is demonstrated by a discussion of the weakness of an “authorial” signal as opposed to the clear distinction among individual works.

Enabling Natural Language Processing and AI Research in Low-Resource Languages: Development and Description of an Assamese UPoS Tagged Dataset

This paper describes in detail the Universal Parts of Speech (UPoS) tagged dataset for the Assamese language. PoS tagged dataset in a language is crucial for experimenting and creating resources for various Natural Language Processing (NLP) and AI research. With the growing usage of Universal Dependency standards, tagged dataset with Universal PoS tags are becoming very much essential for contemporary experiments in the NLP community. NLP research in Assamese, and Indo-Aryan language, is relatively new, and the language is considered a Low Resource language. The dataset of UPoS tagged Assamese text is created with an aim of contributing towards enriching this low resource language for NLP and AI tasks. The size of the dataset is 283506 tokens of Assamese vocabulary, against total 20280 sentences, tagged with 17 standard UPoS tags of core lexical categories. The raw data are taken from an open-source corpus originally tagged with BIS tagset. The original size of 453457 tokens against 29504 sentences, after subjected to data filtering, was reduced to this clean resource of 283506 tokens. Lexical categories mapping is done with linguistic expertise, from BIS to UPoS tagsets. Mapped pattern was used for a first-level conversion of BIS tags to UPoS tags. Linguistic validation is also performed with linguistic experts and inter annotator agreement/disagreements were recorded. Second level validation resulted in deciding on the agreement, producing the final version of the dataset. This Assamese UPoS tagged dataset is the first of its kind with UPoS annotations and shall serve a wider Assamese NLP research community for model training using Machine Learning/Deep Learning Techniques.

A syntax-enhanced parameter generation network for multi-source cross-lingual event extraction

The task of Cross-Lingual Event Extraction (CLEE) aims to transfer event knowledge from rich-resourced languages to low-resourced languages under a zero-shot cross-lingual setting. Despite presenting successful transfer abilities, existing CLEE methods face two limitations. Firstly, a majority of works concentrate on single source CLEE ignoring the benefits of multi-source potentials. Secondly, universal dependency trees have been adopted for learning shared syntactic features across languages while leaving exploration of language-specific features (e.g. word order) insufficient. In this work, we investigate the effectiveness of multi-source CLEE and propose a Syntax-enhanced Parameter Generation Network (SPGN) for the task. SPGN mainly consists of two components, i.e., a parameter generation tree network for determining language-private syntactic knowledge, and an adversarial network for learning language-shared representations. Experiments on widely used ACE2005 and recently released MINION dataset show that our proposed method significantly outperforms existing baseline systems on total 10 languages. Specifically, we observed an improvement of 1.9% in average F1-score over the best-performing baseline on ACE2005 and 3.5% on MINION, highlighting the benefits of leveraging both language-specific and shared features in CLEE. Further analysis confirms the individual effectiveness of each component of the SPGN model, suggesting its potential applicability in other low-resourced language processing tasks.

Toward a Universal Dependencies Treebank of Old English: Representing the Morphological Relatedness of Un-Derivatives

This article deals with one of the aspects involved in the compilation of a treebank of Old English within the framework of Universal Dependencies. More specifically, this study addresses the question of how to account for the remarkable degree of Old English morphological relatedness in a type of treebank designed to stress syntactic similarities across languages. The solution proposed and assessed in this study is the addition of an extra field of annotation for morphological relatedness. The data of this analysis comprise 1106 derivatives attaching the prefix un-. Out of these, there are around 80 morphologically complex nouns, adjectives, verbs, and adverbs whose derivation cannot be described gradually, 33 of which are unique formations or hapax legomena according to the attestations provided by the Dictionary of Old English Corpus. The main conclusion is that the specification of short-distance and long-distance morphological relatedness provides the Old English treebank with a paradigmatic dimension that can be particularly relevant for languages with relatively generalised and transparent derivational morphology.

Cross-linguistic Dependency Length Minimization in scientific language

Abstract We use Universal Dependencies (UD) for the study of cross-linguistic diachronic syntactic complexity reduction. Specifically, we look at whether and how scientific English and German minimize the length of syntactic dependency relations in the Late Modern period (ca. 1650–1900). Our linguistic analysis follows the assumption that over time, scientific discourse cross-linguistically develops towards an increasingly efficient syntactic code by minimizing Dependency Length (DL) as a factor of syntactic complexity. For each language, we analyse a large UD-annotated scientific and general language corpus for comparison. While on a macro level, our analysis suggests that there is an overall diachronic cross-linguistic and cross-register reduction in Average Dependency Length (ADL), on the micro level we find that only scientific language shows a sentence length independent reduction of ADL, while general language shows an overall decrease of ADL due to sentence length reduction. We further analyse the syntactic constructions responsible for this reduction in both languages, showing that both scientific English and German increasingly make use of short, intra-phrasal dependency relations while long dependency relations such as clausal embeddings become rather disfavoured over time.

U-DepPLLaMA: Universal Dependency Parsing via Auto-regressive Large Language Models

This paper investigates the rapidly advancing domain of Large Language Models (LLMs) and their growing potential in various fields. A central focus is the exploration of LLMs, e.g., LLaMA, as powerful tools for modeling and representing linguistic information, especially in the realm of syntax. We aim to evaluate the ability of these models to encode syntactic information, especially when explicitly supplied, through fine-tuning processes. Traditionally, Dependency Parsing has relied on specific techniques and dedicated architectures. Our research shifts this approach, conceptualizing it as a sequence-to-sequence task where Language Models interpret and transform syntax into bracketed structures that reflect dependency graphs. We introduce U-DepPLLaMA (Universal Dependency Parsing via auto-regressive LLMs based on LLaMA), a novel architecture optimized for multilingual, end-to-end Dependency Parsing. Our experimental evaluation, across 50 datasets in 26 languages from the Universal Dependency Treebank, shows that LLMs can be effectively trained for dependency parsing without the need for task-specific architectures. The results are on par with current state-of-the-art methods and demonstrate resilience across varying sentence complexities and lengths.

Identification of Economies of Scale in Regional-Industrial Production Complexes of Russia: Theoretical Foundations and Econometric Estimates

International economic sanctions against Russia are aimed at blocking such a vital phenomenon as the scale effect. In this regard, it is necessary to search for internal reserves for the manifestation of this effect and use it to simultaneously stimulate economic growth and technological progress. The aim of the article is to develop a simple toolkit to identify the scale effect for any industries and regions of the Russian Federation in order to identify the most promising territorial-industry clusters, investments in which can give the maximum return in terms of production efficiency growth. The proposed method was tested by constructing universal econometric dependencies for the agricultural sector in 82 regions of Russia. The obtained quantitative results of the elasticity of labor productivity by production volume made it possible to map the country's agriculture into more and less promising regions. In addition to the effect of scale, the use of an additional factor - the achieved technological level of the region relative to the national average - allowed us to assess the possible structural effect of investment injections into the regional cluster of enterprises, to refine the applied calculations and to select 22 most promising subjects of the Russian Federation, of which 7 regions are capable of becoming the technological driver of the Russian agricultural sector of the economy. These calculations made it possible to draw a map of the most promising industry clusters in Russia. The issue of organizing the borrowing of advanced agrarian technologies within the country - from the advanced regions and enterprises to the lagging ones - is discussed. The inverse relationship between the achieved technological level of enterprises and the value of the scale effect has been established, which indicates the gradual exhaustion of this effect in the process of technological progress.

Data-driven Parsing Evaluation for Child-Parent Interactions

Abstract We present a syntactic dependency treebank for naturalistic child and child-directed spoken English. Our annotations largely follow the guidelines of the Universal Dependencies project (UD [Zeman et al., 2022]), with detailed extensions to lexical and syntactic structures unique to spontaneous spoken language, as opposed to written texts or prepared speech. Compared to existing UD-style spoken treebanks and other dependency corpora of child-parent interactions specifically, our dataset is much larger (44,744 utterances; 233,907 words) and contains data from 10 children covering a wide age range (18–66 months). We conduct thorough dependency parser evaluations using both graph-based and transition-based parsers, trained on three different types of out-of-domain written texts: news, tweets, and learner data. Out-of-domain parsers demonstrate reasonable performance for both child and parent data. In addition, parser performance for child data increases along children’s developmental paths, especially between 18 and 48 months, and gradually approaches the performance for parent data. These results are further validated with in-domain training.