Search

Automatic grammar coaching serves an important purpose of advising on standard grammar varieties while not imposing social pressures or reinforcing established social roles. Such systems already exist but most of them are for English and few of them offer meaningful feedback. Furthermore, they typically rely completely on neural methods and require huge computational resources which most of the world cannot afford. We propose a grammar coaching system for Spanish that relies on (i) a rich linguistic formalism capable of giving informative feedback; and (ii) a faster parsing algorithm which makes using this formalism practical in a real-world application. The approach is feasible for any language for which there is a computerized grammar and is less reliant on expensive and environmentally costly neural methods. We seek to contribute to Greener AI and to address global education challenges by raising the standards of inclusivity and engagement in grammar coaching., Comment: 5 pages, 4 figures, project summary for CEDI-SEPLN Seminar of the Spanish Society for Natural Language Processing at the 7th Spanish Conference on Informatics, June 19-20, 2024, A Coru\~na, Spain

Sentiment analysis is a key technology for companies and institutions to gauge public opinion on products, services or events. However, for large-scale sentiment analysis to be accessible to entities with modest computational resources, it needs to be performed in a resource-efficient way. While some efficient sentiment analysis systems exist, they tend to apply shallow heuristics, which do not take into account syntactic phenomena that can radically change sentiment. Conversely, alternatives that take syntax into account are computationally expensive. The SALSA project, funded by the European Research Council under a Proof-of-Concept Grant, aims to leverage recently-developed fast syntactic parsing techniques to build sentiment analysis systems that are lightweight and efficient, while still providing accuracy and explainability through the explicit use of syntax. We intend our approaches to be the backbone of a working product of interest for SMEs to use in production., Comment: Accepted for publication at SEPLN-CEDI2024: Seminar of the Spanish Society for Natural Language Processing at the 7th Spanish Conference on Informatics

This is a summary of the paper "A Confederacy of Models: a Comprehensive Evaluation of LLMs on Creative Writing", which was published in Findings of EMNLP 2023. We evaluate a range of recent state-of-the-art, instruction-tuned large language models (LLMs) on an English creative writing task, and compare them to human writers. For this purpose, we use a specifically-tailored prompt (based on an epic combat between Ignatius J. Reilly, main character of John Kennedy Toole's "A Confederacy of Dunces", and a pterodactyl) to minimize the risk of training data leakage and force the models to be creative rather than reusing existing stories. The same prompt is presented to LLMs and human writers, and evaluation is performed by humans using a detailed rubric including various aspects like fluency, style, originality or humor. Results show that some state-of-the-art commercial LLMs match or slightly outperform our human writers in most of the evaluated dimensions. Open-source LLMs lag behind. Humans keep a close lead in originality, and only the top three LLMs can handle humor at human-like levels., Comment: Published in the XIX Conference of the Spanish Association for Artificial Intelligence (CAEPIA), 2024. Summary of our paper "A Confederacy of Models: a Comprehensive Evaluation of LLMs on Creative Writing", published in Findings of EMNLP

Sentiment Analysis (SA) is a crucial aspect of Natural Language Processing (NLP), addressing subjective assessments in textual content. Syntactic parsing is useful in SA because explicit syntactic information can improve accuracy while providing explainability, but it tends to be a computational bottleneck in practice due to the slowness of parsing algorithms. This paper addresses said bottleneck by using a SEquence Labeling Syntactic Parser (SELSP) to inject syntax into SA. By treating dependency parsing as a sequence labeling problem, we greatly enhance the speed of syntax-based SA. SELSP is trained and evaluated on a ternary polarity classification task, demonstrating its faster performance and better accuracy in polarity prediction tasks compared to conventional parsers like Stanza and to heuristic approaches that use shallow syntactic rules for SA like VADER. This increased speed and improved accuracy make SELSP particularly appealing to SA practitioners in both research and industry. In addition, we test several sentiment dictionaries on our SELSP to see which one improves the performance in polarity prediction tasks. Moreover, we compare the SELSP with Transformer-based models trained on a 5-label classification task. The results show that dictionaries that capture polarity judgment variation provide better results than dictionaries that ignore polarity judgment variation. Moreover, we show that SELSP is considerably faster than Transformer-based models in polarity prediction tasks.

TripAdvisor reviews and comparable data sources play an important role in many tasks in Natural Language Processing (NLP), providing a data basis for the identification and classification of subjective judgments, such as hotel or restaurant reviews, into positive or negative polarities. This study explores three important factors influencing variation in crowdsourced polarity judgments, focusing on TripAdvisor reviews in Spanish. Three hypotheses are tested: the role of Part Of Speech (POS), the impact of sentiment words such as "tasty", and the influence of neutral words like "ok" on judgment variation. The study's methodology employs one-word titles, demonstrating their efficacy in studying polarity variation of words. Statistical tests on mean equality are performed on word groups of our interest. The results of this study reveal that adjectives in one-word titles tend to result in lower judgment variation compared to other word types or POS. Sentiment words contribute to lower judgment variation as well, emphasizing the significance of sentiment words in research on polarity judgments, and neutral words are associated with higher judgment variation as expected. However, these effects cannot be always reproduced in longer titles, which suggests that longer titles do not represent the best data source for testing the ambiguity of single words due to the influence on word polarity by other words like negation in longer titles. This empirical investigation contributes valuable insights into the factors influencing polarity variation of words, providing a foundation for NLP practitioners that aim to capture and predict polarity judgments in Spanish and for researchers that aim to understand factors influencing judgment variation., Comment: Accepted manuscript to be published in PLoS One

We study incremental constituent parsers to assess their capacity to output trees based on prefix representations alone. Guided by strictly left-to-right generative language models and tree-decoding modules, we build parsers that adhere to a strong definition of incrementality across languages. This builds upon work that asserted incrementality, but that mostly only enforced it on either the encoder or the decoder. Finally, we conduct an analysis against non-incremental and partially incremental models., Comment: Accepted at EACL 2024

We introduce an encoding for parsing as sequence labeling that can represent any projective dependency tree as a sequence of 4-bit labels, one per word. The bits in each word's label represent (1) whether it is a right or left dependent, (2) whether it is the outermost (left/right) dependent of its parent, (3) whether it has any left children and (4) whether it has any right children. We show that this provides an injective mapping from trees to labels that can be encoded and decoded in linear time. We then define a 7-bit extension that represents an extra plane of arcs, extending the coverage to almost full non-projectivity (over 99.9% empirical arc coverage). Results on a set of diverse treebanks show that our 7-bit encoding obtains substantial accuracy gains over the previously best-performing sequence labeling encodings., Comment: Accepted for publication at EMNLP 2023

We evaluate a range of recent LLMs on English creative writing, a challenging and complex task that requires imagination, coherence, and style. We use a difficult, open-ended scenario chosen to avoid training data reuse: an epic narration of a single combat between Ignatius J. Reilly, the protagonist of the Pulitzer Prize-winning novel A Confederacy of Dunces (1980), and a pterodactyl, a prehistoric flying reptile. We ask several LLMs and humans to write such a story and conduct a human evalution involving various criteria such as fluency, coherence, originality, humor, and style. Our results show that some state-of-the-art commercial LLMs match or slightly outperform our writers in most dimensions; whereas open-source LLMs lag behind. Humans retain an edge in creativity, while humor shows a binary divide between LLMs that can handle it comparably to humans and those that fail at it. We discuss the implications and limitations of our study and suggest directions for future research., Comment: Accepted for publication in Findings of EMNLP 2023

Since the popularization of BiLSTMs and Transformer-based bidirectional encoders, state-of-the-art syntactic parsers have lacked incrementality, requiring access to the whole sentence and deviating from human language processing. This paper explores whether fully incremental dependency parsing with modern architectures can be competitive. We build parsers combining strictly left-to-right neural encoders with fully incremental sequence-labeling and transition-based decoders. The results show that fully incremental parsing with modern architectures considerably lags behind bidirectional parsing, noting the challenges of psycholinguistically plausible parsing., Comment: Accepted at IJCNLP-AACL 2023

We present the latest version of the Spanish Resource Grammar (SRG), a grammar of Spanish implemented in the HPSG formalism. Such grammars encode a complex set of hypotheses about syntax making them a resource for empirical testing of linguistic theory. They also encode a strict notion of grammaticality which makes them a resource for natural language processing applications in computer-assisted language learning. This version of the SRG uses the recent version of the Freeling morphological analyzer and is released along with an automatically created, manually verified treebank of 2,291 sentences. We explain the treebanking process, emphasizing how it is different from treebanking with manual annotation and how it contributes to empirically-driven development of syntactic theory. The treebanks' high level of consistency and detail makes them a resource for training high-quality semantic parsers and generally systems that benefit from precise and detailed semantics. Finally, we present the grammar's coverage and overgeneration on 100 sentences from a learner corpus, a new research line related to developing methodologies for robust empirical evaluation of hypotheses in second language acquisition., Comment: 12 pages, 5 figures

We present an approach for assessing how multilingual large language models (LLMs) learn syntax in terms of multi-formalism syntactic structures. We aim to recover constituent and dependency structures by casting parsing as sequence labeling. To do so, we select a few LLMs and study them on 13 diverse UD treebanks for dependency parsing and 10 treebanks for constituent parsing. Our results show that: (i) the framework is consistent across encodings, (ii) pre-trained word vectors do not favor constituency representations of syntax over dependencies, (iii) sub-word tokenization is needed to represent syntax, in contrast to character-based models, and (iv) occurrence of a language in the pretraining data is more important than the amount of task data when recovering syntax from the word vectors., Comment: Accepted at IJCNLP-AACL 2023

We present new supertaggers trained on HPSG-based treebanks. These treebanks feature high-quality annotation based on a well-developed linguistic theory and include diverse and challenging test datasets, beyond the usual WSJ section 23 and Wikipedia data. HPSG supertagging has previously relied on MaxEnt-based models. We use SVM and neural CRF- and BERT-based methods and show that both SVM and neural supertaggers achieve considerably higher accuracy compared to the baseline. Our fine-tuned BERT-based tagger achieves 97.26% accuracy on 1000 sentences from WSJ23 and 93.88% on the completely out-of-domain The Cathedral and the Bazaar (cb)). We conclude that it therefore makes sense to integrate these new supertaggers into modern HPSG parsers, and we also hope that the diverse and difficult datasets we used here will gain more popularity in the field. We contribute the complete dataset reformatted for token classification., Comment: 9 pages, 0 figures

This paper summarizes the results of experimenting with Universal Dependencies (UD) adaptation of an Unsupervised, Compositional and Recursive (UCR) rule-based approach for Sentiment Analysis (SA) submitted to the Shared Task at Rest-Mex 2023 (Team Olga/LyS-SALSA) (within the IberLEF 2023 conference). By using basic syntactic rules such as rules of modification and negation applied on words from sentiment dictionaries, our approach exploits some advantages of an unsupervised method for SA: (1) interpretability and explainability of SA, (2) robustness across datasets, languages and domains and (3) usability by non-experts in NLP. We compare our approach with other unsupervised approaches of SA that in contrast to our UCR rule-based approach use simple heuristic rules to deal with negation and modification. Our results show a considerable improvement over these approaches. We discuss future improvements of our results by using modality features as another shifting rule of polarity and word disambiguation techniques to identify the right sentiment words., Comment: Proceedings of IberLEF 2023, Ja\'en, Spain, 2023

We conduct a quantitative analysis contrasting human-written English news text with comparable large language model (LLM) output from six different LLMs that cover three different families and four sizes in total. Our analysis spans several measurable linguistic dimensions, including morphological, syntactic, psychometric, and sociolinguistic aspects. The results reveal various measurable differences between human and AI-generated texts. Human texts exhibit more scattered sentence length distributions, more variety of vocabulary, a distinct use of dependency and constituent types, shorter constituents, and more optimized dependency distances. Humans tend to exhibit stronger negative emotions (such as fear and disgust) and less joy compared to text generated by LLMs, with the toxicity of these models increasing as their size grows. LLM outputs use more numbers, symbols and auxiliaries (suggesting objective language) than human texts, as well as more pronouns. The sexist bias prevalent in human text is also expressed by LLMs, and even magnified in all of them but one. Differences between LLMs and humans are larger than between LLMs., Comment: Published at Artificial Intelligence Review vol. 57

PoS tags, once taken for granted as a useful resource for syntactic parsing, have become more situational with the popularization of deep learning. Recent work on the impact of PoS tags on graph- and transition-based parsers suggests that they are only useful when tagging accuracy is prohibitively high, or in low-resource scenarios. However, such an analysis is lacking for the emerging sequence labeling parsing paradigm, where it is especially relevant as some models explicitly use PoS tags for encoding and decoding. We undertake a study and uncover some trends. Among them, PoS tags are generally more useful for sequence labeling parsers than for other paradigms, but the impact of their accuracy is highly encoding-dependent, with the PoS-based head-selection encoding being best only when both tagging accuracy and resource availability are high., Comment: Accepted at AACL 2022

We contribute to the discussion on parsing performance in NLP by introducing a measurement that evaluates the differences between the distributions of edge displacement (the directed distance of edges) seen in training and test data. We hypothesize that this measurement will be related to differences observed in parsing performance across treebanks. We motivate this by building upon previous work and then attempt to falsify this hypothesis by using a number of statistical methods. We establish that there is a statistical correlation between this measurement and parsing performance even when controlling for potential covariants. We then use this to establish a sampling technique that gives us an adversarial and complementary split. This gives an idea of the lower and upper bounds of parsing systems for a given treebank in lieu of freshly sampled data. In a broader sense, the methodology presented here can act as a reference for future correlation-based exploratory work in NLP., Comment: This is the final peer-reviewed manuscript accepted for publication in Computational Linguistics. The journal version with the final editorial and typesetting changes is available open-access at https://doi.org/10.1162/coli_a_00440

Treebank selection for parsing evaluation and the spurious effects that might arise from a biased choice have not been explored in detail. This paper studies how evaluating on a single subset of treebanks can lead to weak conclusions. First, we take a few contrasting parsers, and run them on subsets of treebanks proposed in previous work, whose use was justified (or not) on criteria such as typology or data scarcity. Second, we run a large-scale version of this experiment, create vast amounts of random subsets of treebanks, and compare on them many parsers whose scores are available. The results show substantial variability across subsets and that although establishing guidelines for good treebank selection is hard, it is possible to detect potentially harmful strategies., Comment: Accepted at COLING 2022

Machine transliteration, as defined in this paper, is a process of automatically transforming written script of words from a source alphabet into words of another target alphabet within the same language, while preserving their meaning, as well as pronunciation. The main goal of this paper is to present a machine transliteration tool between three common scripts used in low-resource Uzbek language: the old Cyrillic, currently official Latin, and newly announced New Latin alphabets. The tool has been created using a combination of rule-based and fine-tuning approaches. The created tool is available as an open-source Python package, as well as a web-based application including a public API. To our knowledge, this is the first machine transliteration tool that supports the newly announced Latin alphabet of the Uzbek language., Comment: Preprint of a conference paper: The International Conference on Agglutinative Language Technologies as a challenge of Natural Language Processing (ALTNLP)

We propose a morphology-based method for low-resource (LR) dependency parsing. We train a morphological inflector for target LR languages, and apply it to related rich-resource (RR) treebanks to create cross-lingual (x-inflected) treebanks that resemble the target LR language. We use such inflected treebanks to train parsers in zero- (training on x-inflected treebanks) and few-shot (training on x-inflected and target language treebanks) setups. The results show that the method sometimes improves the baselines, but not consistently., Comment: 10 pages, 7 tables, 5 figures. Workshop on Insights from Negative Results in NLP 2022 (co-located with ACL)

Semantic relatedness between words is one of the core concepts in natural language processing, thus making semantic evaluation an important task. In this paper, we present a semantic model evaluation dataset: SimRelUz - a collection of similarity and relatedness scores of word pairs for the low-resource Uzbek language. The dataset consists of more than a thousand pairs of words carefully selected based on their morphological features, occurrence frequency, semantic relation, as well as annotated by eleven native Uzbek speakers from different age groups and gender. We also paid attention to the problem of dealing with rare words and out-of-vocabulary words to thoroughly evaluate the robustness of semantic models., Comment: Final version, published in the proceedings of SIGUL workshop of LREC 2022

This paper addressed the problem of structured sentiment analysis using a bi-affine semantic dependency parser, large pre-trained language models, and publicly available translation models. For the monolingual setup, we considered: (i) training on a single treebank, and (ii) relaxing the setup by training on treebanks coming from different languages that can be adequately processed by cross-lingual language models. For the zero-shot setup and a given target treebank, we relied on: (i) a word-level translation of available treebanks in other languages to get noisy, unlikely-grammatical, but annotated data (we release as much of it as licenses allow), and (ii) merging those translated treebanks to obtain training data. In the post-evaluation phase, we also trained cross-lingual models that simply merged all the English treebanks and did not use word-level translations, and yet obtained better results. According to the official results, we ranked 8th and 9th in the monolingual and cross-lingual setups., Comment: To be published at SemEval-2022

In order to achieve deep natural language understanding, syntactic constituent parsing is a vital step, highly demanded by many artificial intelligence systems to process both text and speech. One of the most recent proposals is the use of standard sequence-to-sequence models to perform constituent parsing as a machine translation task, instead of applying task-specific parsers. While they show a competitive performance, these text-to-parse transducers are still lagging behind classic techniques in terms of accuracy, coverage and speed. To close the gap, we here extend the framework of sequence-to-sequence models for constituent parsing, not only by providing a more powerful neural architecture for improving their performance, but also by enlarging their coverage to handle the most complex syntactic phenomena: discontinuous structures. To that end, we design several novel linearizations that can fully produce discontinuities and, for the first time, we test a sequence-to-sequence model on the main discontinuous benchmarks, obtaining competitive results on par with task-specific discontinuous constituent parsers and achieving state-of-the-art scores on the (discontinuous) English Penn Treebank., Comment: Final peer-reviewed manuscript accepted for publication in Neurocomputing

Dependency distance minimization (DDm) is a well-established principle of word order. It has been predicted theoretically that DDm implies compression, namely the minimization of word lengths. This is a second order prediction because it links a principle with another principle, rather than a principle and a manifestation as in a first order prediction. Here we test that second order prediction with a parallel collection of treebanks controlling for annotation style with Universal Dependencies and Surface-Syntactic Universal Dependencies. To test it, we use a recently introduced score that has many mathematical and statistical advantages with respect to the widely used sum of dependency distances. We find that the prediction is confirmed by the new score when word lengths are measured in phonemes, independently of the annotation style, but not when word lengths are measured in syllables. In contrast, one of the most widely used scores, i.e. the sum of dependency distances, fails to confirm that prediction, showing the weakness of raw dependency distances for research on word order. Finally, our findings expand the theory of natural communication by linking two distinct levels of organization, namely syntax (word order) and word internal structure., Comment: To appear in the proceedings of the SyntaFest 2022 (Sofia, Bulgaria, March 21-25)

We present the system submission from the FASTPARSE team for the EUD Shared Task at IWPT 2021. We engaged in the task last year by focusing on efficiency. This year we have focused on experimenting with new ideas on a limited time budget. Our system is based on splitting the EUD graph into several trees, based on linguistic criteria. We predict these trees using a sequence-labelling parser and combine them into an EUD graph. The results were relatively poor, although not a total disaster and could probably be improved with some polishing of the system's rough edges., Comment: IWPT 2021 Shared Task system description. To be published in proceedings of the 17th International Conference on Parsing Technologies

Dependency parsing is a crucial step towards deep language understanding and, therefore, widely demanded by numerous Natural Language Processing applications. In particular, left-to-right and top-down transition-based algorithms that rely on Pointer Networks are among the most accurate approaches for performing dependency parsing. Additionally, it has been observed for the top-down algorithm that Pointer Networks' sequential decoding can be improved by implementing a hierarchical variant, more adequate to model dependency structures. Considering all this, we develop a bottom-up-oriented Hierarchical Pointer Network for the left-to-right parser and propose two novel transition-based alternatives: an approach that parses a sentence in right-to-left order and a variant that does it from the outside in. We empirically test the proposed neural architecture with the different algorithms on a wide variety of languages, outperforming the original approach in practically all of them and setting new state-of-the-art results on the English and Chinese Penn Treebanks for non-contextualized and BERT-based embeddings., Comment: Final peer-reviewed manuscript accepted for publication in Information Fusion

Discontinuous constituent parsers have always lagged behind continuous approaches in terms of accuracy and speed, as the presence of constituents with discontinuous yield introduces extra complexity to the task. However, a discontinuous tree can be converted into a continuous variant by reordering tokens. Based on that, we propose to reduce discontinuous parsing to a continuous problem, which can then be directly solved by any off-the-shelf continuous parser. To that end, we develop a Pointer Network capable of accurately generating the continuous token arrangement for a given input sentence and define a bijective function to recover the original order. Experiments on the main benchmarks with two continuous parsers prove that our approach is on par in accuracy with purely discontinuous state-of-the-art algorithms, but considerably faster., Comment: Camera-ready version to appear at proceedings of EMNLP 2021. 9 pages (incl. appendices)

We present an error analysis of neural UPOS taggers to evaluate why using gold standard tags has such a large positive contribution to parsing performance while using predicted UPOS tags either harms performance or offers a negligible improvement. We evaluate what neural dependency parsers implicitly learn about word types and how this relates to the errors taggers make to explain the minimal impact using predicted tags has on parsers. We also present a short analysis on what contexts result in reductions in tagging performance. We then mask UPOS tags based on errors made by taggers to tease away the contribution of UPOS tags which taggers succeed and fail to classify correctly and the impact of tagging errors., Comment: Due to be published in the proceedings of the 23rd Nordic Conference on Computational Linguistics (NoDaLiDa 2021). Previously rejected at the 2021 Conference of the European Chapter of the Association for Computational Linguistics (EACL 2021)

This paper presents a monolingual BERT model for Galician. We follow the recent trend that shows that it is feasible to build robust monolingual BERT models even for relatively low-resource languages, while performing better than the well-known official multilingual BERT (mBERT). More particularly, we release two monolingual Galician BERT models, built using 6 and 12 transformer layers, respectively; trained with limited resources (~45 million tokens on a single GPU of 24GB). We then provide an exhaustive evaluation on a number of tasks such as POS-tagging, dependency parsing and named entity recognition. For this purpose, all these tasks are cast in a pure sequence labeling setup in order to run BERT without the need to include any additional layers on top of it (we only use an output classification layer to map the contextualized representations into the predicted label). The experiments show that our models, especially the 12-layer one, outperform the results of mBERT in most tasks., Comment: Accepted in the journal Procesamiento del Lenguaje Natural

We present a bracketing-based encoding that can be used to represent any 2-planar dependency tree over a sentence of length n as a sequence of n labels, hence providing almost total coverage of crossing arcs in sequence labeling parsing. First, we show that existing bracketing encodings for parsing as labeling can only handle a very mild extension of projective trees. Second, we overcome this limitation by taking into account the well-known property of 2-planarity, which is present in the vast majority of dependency syntactic structures in treebanks, i.e., the arcs of a dependency tree can be split into two planes such that arcs in a given plane do not cross. We take advantage of this property to design a method that balances the brackets and that encodes the arcs belonging to each of those planes, allowing for almost unrestricted non-projectivity (round 99.9% coverage) in sequence labeling parsing. The experiments show that our linearizations improve over the accuracy of the original bracketing encoding in highly non-projective treebanks (on average by 0.4 LAS), while achieving a similar speed. Also, they are especially suitable when PoS tags are not used as input parameters to the models., Comment: COLING2020 (long papers), 13 pages (incl. appendix) with corrected parsing speeds for Danish and Gothic

We define a mapping from transition-based parsing algorithms that read sentences from left to right to sequence labeling encodings of syntactic trees. This not only establishes a theoretical relation between transition-based parsing and sequence-labeling parsing, but also provides a method to obtain new encodings for fast and simple sequence labeling parsing from the many existing transition-based parsers for different formalisms. Applying it to dependency parsing, we implement sequence labeling versions of four algorithms, showing that they are learnable and obtain comparable performance to existing encodings., Comment: Camera-ready version (final peer-reviewed manuscript) to appear at proceedings of COLING 2020. 18 pages (incl. appendices)

We present an analysis on the effect UPOS accuracy has on parsing performance. Results suggest that leveraging UPOS tags as features for neural parsers requires a prohibitively high tagging accuracy and that the use of gold tags offers a non-linear increase in performance, suggesting some sort of exceptionality. We also investigate what aspects of predicted UPOS tags impact parsing accuracy the most, highlighting some potentially meaningful linguistic facets of the problem., Comment: To be published in proceedings of the 24th SIGNLL Conference on Computational Natural Language Learning (CoNLL). Be aware of long appendix: please don't print all 28 pages

This paper reduces discontinuous parsing to sequence labeling. It first shows that existing reductions for constituent parsing as labeling do not support discontinuities. Second, it fills this gap and proposes to encode tree discontinuities as nearly ordered permutations of the input sequence. Third, it studies whether such discontinuous representations are learnable. The experiments show that despite the architectural simplicity, under the right representation, the models are fast and accurate., Comment: To appear in EMNLP 2020

We propose a transition-based approach that, by training a single model, can efficiently parse any input sentence with both constituent and dependency trees, supporting both continuous/projective and discontinuous/non-projective syntactic structures. To that end, we develop a Pointer Network architecture with two separate task-specific decoders and a common encoder, and follow a multitask learning strategy to jointly train them. The resulting quadratic system, not only becomes the first parser that can jointly produce both unrestricted constituent and dependency trees from a single model, but also proves that both syntactic formalisms can benefit from each other during training, achieving state-of-the-art accuracies in several widely-used benchmarks such as the continuous English and Chinese Penn Treebanks, as well as the discontinuous German NEGRA and TIGER datasets., Comment: Final peer-reviewed manuscript accepted for publication in Knowledge-Based Systems

It is often stated that human languages, as other biological systems, are shaped by cost-cutting pressures but, to what extent? Attempts to quantify the degree of optimality of languages by means of an optimality score have been scarce and focused mostly on English. Here we recast the problem of the optimality of the word order of a sentence as an optimization problem on a spatial network where the vertices are words, arcs indicate syntactic dependencies and the space is defined by the linear order of the words in the sentence. We introduce a new score to quantify the cognitive pressure to reduce the distance between linked words in a sentence. The analysis of sentences from 93 languages representing 19 linguistic families reveals that half of languages are optimized to a 70% or more. The score indicates that distances are not significantly reduced in a few languages and confirms two theoretical predictions, i.e. that longer sentences are more optimized and that distances are more likely to be longer than expected by chance in short sentences. We present a new hierarchical ranking of languages by their degree of optimization. The new score has implications for various fields of language research (dependency linguistics, typology, historical linguistics, clinical linguistics and cognitive science). Finally, the principles behind the design of the score have implications for network science., Comment: A new version incorporating reviewer comments

A fundamental problem in network science is the normalization of the topological or physical distance between vertices, that requires understanding the range of variation of the unnormalized distances. Here we investigate the limits of the variation of the physical distance in linear arrangements of the vertices of trees. In particular, we investigate various problems on the sum of edge lengths in trees of a fixed size: the minimum and the maximum value of the sum for specific trees, the minimum and the maximum in classes of trees (bistar trees and caterpillar trees) and finally the minimum and the maximum for any tree. We establish some foundations for research on optimality scores for spatial networks in one dimension., Comment: Title changed at proof stage

The carbon footprint of natural language processing research has been increasing in recent years due to its reliance on large and inefficient neural network implementations. Distillation is a network compression technique which attempts to impart knowledge from a large model to a smaller one. We use teacher-student distillation to improve the efficiency of the Biaffine dependency parser which obtains state-of-the-art performance with respect to accuracy and parsing speed (Dozat and Manning, 2017). When distilling to 20\% of the original model's trainable parameters, we only observe an average decrease of $\sim$1 point for both UAS and LAS across a number of diverse Universal Dependency treebanks while being 2.30x (1.19x) faster than the baseline model on CPU (GPU) at inference time. We also observe a small increase in performance when compressing to 80\% for some treebanks. Finally, through distillation we attain a parser which is not only faster but also more accurate than the fastest modern parser on the Penn Treebank., Comment: To be published in proceedings of the 16th International Conference on Parsing Technologies. Earlier versions were rejected at the 58th Annual Conference of the Association for Computational Linguistics and 8th International Conference on Learning Representations

We present the system submission from the FASTPARSE team for the EUD Shared Task at IWPT 2020. We engaged with the task by focusing on efficiency. For this we considered training costs and inference efficiency. Our models are a combination of distilled neural dependency parsers and a rule-based system that projects UD trees into EUD graphs. We obtained an average ELAS of 74.04 for our official submission, ranking 4th overall., Comment: To published in the proceedings of the IWPT 2020 Shared Task on Parsing into Enhanced Universal Dependencies

Sequence-to-sequence constituent parsing requires a linearization to represent trees as sequences. Top-down tree linearizations, which can be based on brackets or shift-reduce actions, have achieved the best accuracy to date. In this paper, we show that these results can be improved by using an in-order linearization instead. Based on this observation, we implement an enriched in-order shift-reduce linearization inspired by Vinyals et al. (2015)'s approach, achieving the best accuracy to date on the English PTB dataset among fully-supervised single-model sequence-to-sequence constituent parsers. Finally, we apply deterministic attention mechanisms to match the speed of state-of-the-art transition-based parsers, thus showing that sequence-to-sequence models can match them, not only in accuracy, but also in speed., Comment: Proceedings of ACL 2020. 8 pages

Transition-based parsers implemented with Pointer Networks have become the new state of the art in dependency parsing, excelling in producing labelled syntactic trees and outperforming graph-based models in this task. In order to further test the capabilities of these powerful neural networks on a harder NLP problem, we propose a transition system that, thanks to Pointer Networks, can straightforwardly produce labelled directed acyclic graphs and perform semantic dependency parsing. In addition, we enhance our approach with deep contextualized word embeddings extracted from BERT. The resulting system not only outperforms all existing transition-based models, but also matches the best fully-supervised accuracy to date on the SemEval 2015 Task 18 English datasets among previous state-of-the-art graph-based parsers., Comment: Proceedings of ACL 2020. 12 pages

There has been an increasing interest in learning cross-lingual word embeddings to transfer knowledge obtained from a resource-rich language, such as English, to lower-resource languages for which annotated data is scarce, such as Turkish, Russian, and many others. In this paper, we present the first viability study of established techniques to align monolingual embedding spaces for Turkish, Uzbek, Azeri, Kazakh and Kyrgyz, members of the Turkic family which is heavily affected by the low-resource constraint. Those techniques are known to require little explicit supervision, mainly in the form of bilingual dictionaries, hence being easily adaptable to different domains, including low-resource ones. We obtain new bilingual dictionaries and new word embeddings for these languages and show the steps for obtaining cross-lingual word embeddings using state-of-the-art techniques. Then, we evaluate the results using the bilingual dictionary induction task. Our experiments confirm that the obtained bilingual dictionaries outperform previously-available ones, and that word embeddings from a low-resource language can benefit from resource-rich closely-related languages when they are aligned together. Furthermore, evaluation on an extrinsic task (Sentiment analysis on Uzbek) proves that monolingual word embeddings can, although slightly, benefit from cross-lingual alignments., Comment: Final version, published in the proceedings of LREC 2020

A wide variety of transition-based algorithms are currently used for dependency parsers. Empirical studies have shown that performance varies across different treebanks in such a way that one algorithm outperforms another on one treebank and the reverse is true for a different treebank. There is often no discernible reason for what causes one algorithm to be more suitable for a certain treebank and less so for another. In this paper we shed some light on this by introducing the concept of an algorithm's inherent dependency displacement distribution. This characterises the bias of the algorithm in terms of dependency displacement, which quantify both distance and direction of syntactic relations. We show that the similarity of an algorithm's inherent distribution to a treebank's displacement distribution is clearly correlated to the algorithm's parsing performance on that treebank, specifically with highly significant and substantial correlations for the predominant sentence lengths in Universal Dependency treebanks. We also obtain results which show a more discrete analysis of dependency displacement does not result in any meaningful correlations., Comment: To be published in proceedings of the 12th Language Resources and Evaluation Conference. Earlier versions were rejected at the 57th Annual Conference of the Association for Computational Linguistics and the SIGNLL Conference on Computational Natural Language Learning, 2019

One of the most complex syntactic representations used in computational linguistics and NLP are discontinuous constituent trees, crucial for representing all grammatical phenomena of languages such as German. Recent advances in dependency parsing have shown that Pointer Networks excel in efficiently parsing syntactic relations between words in a sentence. This kind of sequence-to-sequence models achieve outstanding accuracies in building non-projective dependency trees, but its potential has not been proved yet on a more difficult task. We propose a novel neural network architecture that, by means of Pointer Networks, is able to generate the most accurate discontinuous constituent representations to date, even without the need of Part-of-Speech tagging information. To do so, we internally model discontinuous constituent structures as augmented non-projective dependency structures. The proposed approach achieves state-of-the-art results on the two widely-used NEGRA and TIGER benchmarks, outperforming previous work by a wide margin., Comment: Proceedings of AAAI 2020. 8 pages

Recent analyses suggest that encoders pretrained for language modeling capture certain morpho-syntactic structure. However, probing frameworks for word vectors still do not report results on standard setups such as constituent and dependency parsing. This paper addresses this problem and does full parsing (on English) relying only on pretraining architectures -- and no decoding. We first cast constituent and dependency parsing as sequence tagging. We then use a single feed-forward layer to directly map word vectors to labels that encode a linearized tree. This is used to: (i) see how far we can reach on syntax modelling with just pretrained encoders, and (ii) shed some light about the syntax-sensitivity of different word vectors (by freezing the weights of the pretraining network during training). For evaluation, we use bracketing F1-score and LAS, and analyze in-depth differences across representations for span lengths and dependency displacements. The overall results surpass existing sequence tagging parsers on the PTB (93.5%) and end-to-end EN-EWT UD (78.8%)., Comment: AAAI 2020 - The Thirty-Fourth AAAI Conference on Artificial Intelligence

Research on word embeddings has mainly focused on improving their performance on standard corpora, disregarding the difficulties posed by noisy texts in the form of tweets and other types of non-standard writing from social media. In this work, we propose a simple extension to the skipgram model in which we introduce the concept of bridge-words, which are artificial words added to the model to strengthen the similarity between standard words and their noisy variants. Our new embeddings outperform baseline models on noisy texts on a wide range of evaluation tasks, both intrinsic and extrinsic, while retaining a good performance on standard texts. To the best of our knowledge, this is the first explicit approach at dealing with this type of noisy texts at the word embedding level that goes beyond the support for out-of-vocabulary words., Comment: 15 pages, 2 figures, 5 tables

We explore whether it is possible to leverage eye-tracking data in an RNN dependency parser (for English) when such information is only available during training, i.e., no aggregated or token-level gaze features are used at inference time. To do so, we train a multitask learning model that parses sentences as sequence labeling and leverages gaze features as auxiliary tasks. Our method also learns to train from disjoint datasets, i.e. it can be used to test whether already collected gaze features are useful to improve the performance on new non-gazed annotated treebanks. Accuracy gains are modest but positive, showing the feasibility of the approach. It can serve as a first step towards architectures that can better leverage eye-tracking data or other complementary information available only for training sentences, possibly leading to improvements in syntactic parsing., Comment: Camera-ready version to appear at EMNLP 2019 (final peer-reviewed manuscript). 8 pages (incl. appendix)

The ability to produce and understand an unlimited number of different sentences is a hallmark of human language. Linguists have sought to define the essence of this generative capacity using formal grammars that describe the syntactic dependencies between constituents, independent of the computational limitations of the human brain. Here, we evaluate this independence assumption by sampling sentences uniformly from the space of possible syntactic structures. We find that the average dependency distance between syntactically related words, a proxy for memory limitations, is less than expected by chance in a collection of state-of-the-art classes of dependency grammars. Our findings indicate that memory limitations have permeated grammatical descriptions, suggesting that it may be impossible to build a parsimonious theory of human linguistic productivity independent of non-linguistic cognitive constraints., Comment: Improved with reviewer feedback once again. In press in Glottometrics