Skip to main content

TECNOLOGIAS DE LA LENGUA PARA ANALISIS DE OPINIONES EN REDES SOCIALES: DEL TEXTO AL MICROTEXTO

FFI2014-51978-C2-2-R

•
Funding agency name Ministerio de Economía y Competitividad
Funding agency acronym MINECO
Program Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Call Proyectos I+D+i Retos de Investigación
Call year 2014
State Plan 2013-2016 Plan Estatal de Investigación Científica y Técnica y de Innovación
Management unit Secretaría de Estado de Educación, Formación Profesional y Universidades
Beneficiary centre UNIVERSIDADE DA CORUÑA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 18
Encontrada(s) 1 página(s)

Investigo. Repositorio Institucional de la Universidade de Vigo
Publication Contribución de congreso (ConferenceOutput)

An efficient dynamic oracle for unrestricted non-projective parsing

  • Gómez Rodríguez, Carlos
  • Fernández González, Daniel
We define a dynamic oracle for the Covington non-projective dependency parser. This is not only the first dynamic oracle that supports arbitrary non-projectivity, but also considerably more efficient (O(n)) than the only existing oracle with restricted non-projectivity support. Experiments show that training with the dynamic oracle significantly improves parsing accuracy over the static oracle baseline on a wide range of treebanks., Ministerio de Economía y Competitividad | Ref. FFI2014-51978-C2-1-R, Ministerio de Economía y Competitividad | Ref. FFI2014-51978-C2-2-R, Xunta de Galicia | Ref. R2014/034
Investigo. Repositorio Institucional de la Universidade de Vigo
Publication Artículo científico (JournalArticle)

On the performance of phonetic algorithms in microtext normalization

  • Doval Mosquera, Yerai
  • Vilares Ferro, Manuel
  • Vilares Ferro, Jesús
User–generated content published on microblogging social networks constitutes a priceless source of information. However, microtexts usually deviate from the standard lexical and grammatical rules of the language, thus making its processing by traditional intelligent systems very difficult. As an answer, microtext normalization consists in transforming those non–standard microtexts into standard well–written texts as a preprocessing step, allowing traditional approaches to continue with their usual processing. Given the importance of phonetic phenomena in non–standard text formation, an essential element of the knowledge base of a normalizer would be the phonetic rules that encode these phenomena, which can be found in the so–called phonetic algorithms. In this work we experiment with a wide range of phonetic algorithms for the English language. The aim of this study is to determine the best phonetic algorithms within the context of candidate generation for microtext normalization. In other words, we intend to find those algorithms that taking as input non–standard terms to be normalized allow us to obtain as output the smallest possible sets of normalization candidates which still contain the corresponding target standard words. As it will be stated, the choice of the phonetic algorithm will depend heavily on the capabilities of the candidate selection mechanism which we usually find at the end of a microtext normalization pipeline. The faster it can make the right choices among big enough sets of candidates, the more we can sacrifice on the precision of the phonetic algorithms in favour of coverage in order to increase the overall performance of the normalization system, Agencia Estatal de Investigación | Ref. TIN2017-85160-C2-1-R, Agencia Estatal de Investigación | Ref. TIN2017-85160-C2-2-R, Ministerio de Economía y Competitividad | Ref. FFI2014-51978-C2-1-R, Ministerio de Economía y Competitividad | Ref. FFI2014-51978-C2-2-R, Xunta de Galicia | Ref. ED431D-2017/12, Xunta de Galicia | Ref. ED431B2017/01, Xunta de Galicia | Ref. ED431D R2016/046, Ministerio de Economía y Competitividad | Ref. BES-2015-073768
RUA. Repositorio Institucional de la Universidad de Alicante
Publication Otros (Other)

Segmentación de palabras en español mediante modelos del lenguaje basados en redes neuronales, Spanish word segmentation through neural language models

  • Doval, Yerai
  • Gómez Rodríguez, Carlos
  • Vilares Ferro, Jesús
En las plataformas de microblogging abundan ciertos tokens especiales como los hashtags o las menciones en los que un grupo de palabras se escriben juntas sin espaciado entre ellas; p.ej.: #añobisiesto o @ryanreynoldsnet. Debido a la forma en que se escriben este tipo de textos, este fenómeno de ensamblado de palabras puede aparecer junto a su opuesto, la segmentación de palabras, afectando a cualquier elemento del texto y dificultando su análisis. En este trabajo se muestra un enfoque algorítmico que utiliza como base un modelo del lenguaje - en nuestro caso concreto uno basado en redes neuronales - para resolver el problema de la segmentación y ensamblado de palabras, en el que se trata de recuperar el espaciado estándar de las palabras que han sufrido alguna de estas transformaciones añadiendo o quitando espacios donde corresponda. Los resultados obtenidos son prometedores e indican que tras un mayor refinamiento del modelo del lenguaje se podrá sobrepasar al estado del arte.
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

Grounding the Semantics of Part-of-Day Nouns Worldwide using Twitter

  • Vilares, David
  • Gómez-Rodríguez, Carlos
Held 6 June 2018, New Orleans, Louisiana, USA., [Absctract]: The usage of part-of-day nouns, such as ‘night’, and their time-specific greetings (‘good night’), varies across languages and cultures. We show the possibilities that Twitter offers for studying the semantics of these terms and its variability between countries. We mine a worldwide sample of multilingual tweets with temporal greetings, and study how their frequencies vary in relation with local time. The results provide insights into the semantics of these temporal expressions and the cultural and sociological factors influencing their usage., DV and CGR receive funding from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, grant agreement No 714150), from the TELEPARESUDC project (FFI2014-51978-C2-2-R) and the ANSWER-ASAP project (TIN2017-85160-C2-1- R) from MINECO, and from Xunta de Galicia (ED431B 2017/01)., Xunta de Galicia; ED431B 2017/01
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

Global Transition-based Non-projective Dependency Parsing

  • Fernández-González, Daniel
  • Shi, Tianze
  • Lee, Lillian
56th Annual Meeting of the Association for Computational Linguistics, ACL 2018. Melbourne15 July - 20 July 2018., [Absctract]: Shi, Huang, and Lee (2017a) obtained state-of-the-art results for English and Chinese dependency parsing by combining dynamic-programming implementations of transition-based dependency parsers with a minimal set of bidirectional LSTM features. However, their results were limited to projective parsing. In this paper, we extend their approach to support non-projectivity by providing the first practical implementation of the MH4 algorithm, an Opn4q mildly non projective dynamic-programming parser with very high coverage on non-projective treebanks. To make MH4 compatible with minimal transition-based feature sets, we introduce a transition-based interpretation of it in which parser items are mapped to sequences of transitions. We thus obtain the first implementation of global decoding for non-projective transition-based parsing, and demonstrate empirically that it is more effective than its projective counterpart in parsing a number of highly non-projective languages., We thank the three anonymous reviewers for their helpful comments. CG has received funding from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, grant agreement No 714150), from the TELEPARESUDC project (FFI2014-51978-C2-2-R) and the ANSWER-ASAP project (TIN2017-85160-C2-1-R) from MINECO, and from Xunta de Galicia (ED431B 2017/01). TS and LL were supported in part by a Google Focused Research Grant to Cornell University. LL was also supported in part by NSF grant SES-1741441. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation or other sponsors ., Xunta de Galicia; ED431B 2017/01, Estados Unidos. National Science Foundation; SES-1741441
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

Harry Potter and the Action Prediction Challenge from Natural Language

  • Vilares, David
  • Gómez-Rodríguez, Carlos
The 17th Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT 2019) was held in Minneapolis from June 2nd to June 7th, 2019., [Absctract]: We explore the challenge of action prediction from textual descriptions of scenes, a testbed to approximate whether text inference can be used to predict upcoming actions. As a case of study, we consider the world of the Harry Potter fantasy novels and inferring what spell will be cast next given a fragment of a story. Spells act as keywords that abstract actions (e.g. ‘Alohomora’ to open a door) and denote a response to the environment. This idea is used to automatically build HPAC, a corpus containing 82,836 samples and 85 actions. We then evaluate different baselines. Among the tested models, an LSTM-based approach obtains the best performance for frequent actions and large scene descriptions, but approaches such as logistic regression behave well on infrequent actions., This work has received support from the TELEPARES-UDC project (FFI2014-51978-C2- 2-R) and the ANSWER-ASAP project (TIN2017- 85160-C2-1-R) from MINECO, and from Xunta de Galicia (ED431B 2017/01), and from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, grant agreement No 714150), Xunta de Galicia; ED431B 2017/01
RUC. Repositorio da Universidade da Coruña
Publication Artículo científico (JournalArticle)

On the Degree in Categories of Complexes of Fixed Size

  • Chaio, Claudia
  • Pratti, Isabel
  • Souto Salorio, María José
This version of the article has been accepted for publication, after peer review and is subject to Springer Nature’s AM terms of use, but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: https://doi.org/10.1007/s10485-019-09557-x., [Abstract]: We consider Λ an artin algebra and n ≥ 2. We study how to compute the left and right degrees of irreducible morphisms between complexes in a generalized standard Auslander-Reiten component of Cn(proj Λ) with length. We give conditions under which the kernel and the cokernel of irreducible morphisms between complexes in Cn(proj Λ) belong to such a category. For a finite dimensional hereditary algebra H over an algebraically closed field, we determine when an irreducible morphism has finite left (or right) degree and we give a characterization, depending on the degrees of certain irreducible morphisms, under which Cn(proj H) is of finite type., The first and second authors thankfully acknowledge partial support from CONICET and EXA558/14 from Universidad Nacional de Mar del Plata, Argentina. The third author thankfully acknowledge support from Ministerio Español de Economía y Competitividad and FEDER (FFI2014-51978-C2-2-R). The first author is a researcher from CONICET., Argentina. Universidad Nacional de Mar del Plata; EXA558/14, Cn(proj Λ) H n ≥ 2.
RUC. Repositorio da Universidade da Coruña
Publication Artículo científico (JournalArticle)

On the performance of phonetic algorithms in microtext normalization

  • Doval, Yerai
  • Vilares Ferro, Manuel
  • Vilares, Jesús
© 2018. This manuscript version is made available under the CC-BY-NC-ND 4.0 license https://creativecommons.org/licenses/by-nc-nd/4.0/. This version of the article: Doval, Y., Vilares, M. and Vilares, J. (2018) ‘On the performance of phonetic algorithms in microtext normalization’ has been accepted for publication in: Expert Systems with Applications, 113, pp. 213–222. The Version of Record is available online at: https://doi.org/10.1016/j.eswa.2018.07.016, [Abstract]: User–generated content published on microblogging social networks constitutes a priceless source of information. However, microtexts usually deviate from the standard lexical and grammatical rules of the language, thus making its processing by traditional intelligent systems very difficult. As an answer, microtext normalization consists in transforming those non–standard microtexts into standard well–written texts as a preprocessing step, allowing traditional approaches to continue with their usual processing. Given the importance of phonetic phenomena in non–standard text formation, an essential element of the knowledge base of a normalizer would be the phonetic rules that encode these phenomena, which can be found in the so–called phonetic algorithms. In this work we experiment with a wide range of phonetic algorithms for the English language. The aim of this study is to determine the best phonetic algorithms within the context of candidate generation for microtext normalization. In other words, we intend to find those algorithms that taking as input non–standard terms to be normalized allow us to obtain as output the smallest possible sets of normalization candidates which still contain the corresponding target standard words. As it will be stated, the choice of the phonetic algorithm will depend heavily on the capabilities of the candidate selection mechanism which we usually find at the end of a microtext normalization pipeline. The faster it can make the right choices among big enough sets of candidates, the more we can sacrifice on the precision of the phonetic algorithms in favour of coverage in order to increase the overall performance of the normalization system., This research has been partially funded by the Spanish Ministry of Economy, Industry and Competitiveness (MINECO) through projects TIN2017-85160-C2-1-R, TIN2017-85160-C2-2-R, FFI2014-51978-C2-1-R and FFI2014-51978-C2-2-R, and by the Autonomous Government of Galicia through projects ED431D-2017/12, ED431B-2017/01 and ED431D R2016/046. Moreover, Yerai Doval is funded by the Spanish State Secretariat for Research, Development and Innovation (which belongs to MINECO) and by the European Social Fund (ESF) under a FPI fellowship (BES-2015-073768) associated to project FFI2014-51978-C2-1-R., Xunta de Galicia; ED431D-2017/12, Xunta de Galicia; ED431B-2017/01, Xunta de Galicia; ED431D R2016/046
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

Detecting Perspectives in Political Debates

  • Vilares, David
  • He, Yulan
[Abstract]: We explore how to detect people’s perspectives that occupy a certain proposition. We propose a Bayesian modelling approach where topics (or propositions) and their associated perspectives (or viewpoints) are modeled as latent variables. Words associated with topics or perspectives follow different generative routes. Based on the extracted perspectives, we can extract the top associated sentences from text to generate a succinct summary which allows a quick glimpse of the main viewpoints in a document. The model is evaluated on debates from the House of Commons of the UK Parliament, revealing perspectives from the debates without the use of labelled data and obtaining better results than previous related solutions under a variety of evaluations., We thank Charles Marshall for crawling the HCD data. DV was funded by MECD (FPU 13/01180), MINECO (FFI2014-51978-C2-2-R) and Inditex-UDC grants for research stays. YH is partly funded by the Natural Science Foundation of China (61528302)., China. Natural Science Foundation of China; 61528302
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

Left-to-Right Dependency Parsing with Pointer Networks

  • Fernández-González, Daniel
  • Gómez-Rodríguez, Carlos
Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers), [Abstract]: We propose a novel transition-based algorithm that straightforwardly parses sentences from left to right by building n attachments, with n being the length of the input sentence. Similarly to the recent stack-pointer parser by Ma et al. (2018), we use the pointer network framework that, given a word, can directly point to a position from the sentence. However, our left-to-right approach is simpler than the original top-down stack-pointer parser (not requiring a stack) and reduces transition sequence length in half, from 2n-1 actions to n. This results in a quadratic non-projective parser that runs twice as fast as the original while achieving the best accuracy to date on the English PTB dataset (96.04% UAS, 94.43% LAS) among fully-supervised single-model dependency parsers, and improves over the former top-down transition system in the majority of languages tested., This work has received funding from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, grant agreement No 714150), from MINECO (FFI2014-51978-C2-2-R, TIN2017-85160-C2-1-R) and from Xunta de Galicia (ED431B 2017/01)., Xunta de Galicia; ED431B 2017/01
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

Viable Dependency Parsing as Sequence Labeling

  • Strzyz, Michalina
  • Vilares, David
  • Gómez-Rodríguez, Carlos
[Abstract]: We recast dependency parsing as a sequence labeling problem, exploring several encodings of dependency trees as labels. While dependency parsing by means of sequence labeling had been attempted in existing work, results suggested that the technique was impractical. We show instead that with a conventional BILSTM-based model it is possible to obtain fast and accurate parsers. These parsers are conceptually simple, not needing traditional parsing algorithms or auxiliary structures. However, experiments on the PTB and a sample of UD treebanks show that they provide a good speed-accuracy tradeoff, with results competitive with more complex approaches., This work has received funding from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, grant agreement No 714150), from the TELEPARESUDC project (FFI2014-51978-C2-2-R) and the ANSWER-ASAP project (TIN2017-85160-C2-1-R) from MINECO, and from Xunta de Galicia (ED431B 2017/01). We gratefully acknowledge NVIDIA Corporation for the donation of a GTX Titan X GPU., Xunta de Galicia; ED431B 2017/01
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

Seguimiento y análisis automático de contenidos en redes sociales

  • Alonso, Miguel A.
  • Gómez-Rodríguez, Carlos
  • Vilares, David
  • Doval, Yerai
  • Vilares, Jesús
[Abstract]: La Minería de Opiniones es la disciplina que aborda el tratamiento automático de las opiniones contenidas en un texto. Permite, por ejemplo, determinar si en un texto se está opinando o no, o si la polaridad o sentimiento que se expresa en el mismo es positiva, negativa o mixta. También permite la extracción automática de características, lo que posibilita conocer la percepción que los autores tienen sobre aspectos concretos de un tema determinado. Este trabajo, tras realizar una introducción a dicho ámbito, presenta una aproximación propia al mismo, la cual destaca por emplear información sintáctica, así como por estar especialmente adaptada a uno de los contextos de trabajo más complicados, Twitter. Dicha tecnología es fácilmente aplicable a tareas de inteligencia., Investigación financiada por Ministerio de Economía y Competitividad (FFI2014-51978-C2-2-R), Xunta de Galicia (R2014/034) y Ministerio de Educación, Cultura y Deporte (FPU13/01180)., Xunta de Galicia; R2014/034
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

A Transition-Based Algorithm for Unrestricted AMR Parsing

  • Vilares, David
  • Gómez-Rodríguez, Carlos
The Conference was held in New Orleans, June 1 to June 6, 2018 at the Hyatt Regency New Orleans hotel., [Absctract]: Non-projective parsing can be useful to handle cycles and reentrancy in AMR graphs. We explore this idea and introduce a greedy left-to-right non-projective transition-based parser. At each parsing configuration, an oracle decides whether to create a concept or whether to connect a pair of existing concepts. The algorithm handles reentrancy and arbitrary cycles natively, i.e. within the transition system itself. The model is evaluated on the LDC2015E86 corpus, obtaining results close to the state of the art, including a Smatch of 64%, and showing good behavior on reentrant edges., This work is funded from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, grant agreement No 714150), from the TELEPARES-UDC project (FFI2014- 51978-C2-2-R) and the ANSWER-ASAP project (TIN2017-85160-C2-1-R) from MINECO, and from Xunta de Galicia (ED431B 2017/01). We gratefully acknowledge NVIDIA Corporation for the donation of a GTX Titan X GPU., Xunta de Galicia; ED431B 2017/01
RUC. Repositorio da Universidade da Coruña
Publication Artículo científico (JournalArticle)

Comparing neural- and N-gram-based language models for word segmentation

  • Doval, Yerai
  • Gómez-Rodríguez, Carlos
[Abstract]: Word segmentation is the task of inserting or deleting word boundary characters in order to separate character sequences that correspond to words in some language. In this article we propose an approach based on a beam search algorithm and a language model working at the byte/character level, the latter component implemented either as an n-gram model or a recurrent neural network. The resulting system analyzes the text input with no word boundaries one token at a time, which can be a character or a byte, and uses the information gathered by the language model to determine if a boundary must be placed in the current position or not. Our aim is to use this system in a preprocessing step for a microtext normalization system. This means that it needs to effectively cope with the data sparsity present on this kind of texts. We also strove to surpass the performance of two readily available word segmentation systems: The well-known and accessible Word Breaker by Microsoft, and the Python module WordSegment by Grant Jenks. The results show that we have met our objectives, and we hope to continue to improve both the precision and the efficiency of our system in the future. © 2018 The Authors. Journal of the Association for Information Science and Technology published by Wiley Periodicals, Inc. on behalf of ASIS&T., This research received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (grant agreement no. 714150-FASTPARSE). It was partially funded by the Spanish Ministry of Economy and Competitiveness (MINECO) through projects FFI2014-51978-C2-1-R and FFI2014-51978-C2-2-R, and by the Autonomous Government of Galicia through both the Galician Network for Lexicography-RELEX (ED431D R2016/046) and Grant ED431B-2017/01. Moreover, Yerai Doval is funded by the Spanish State Secretariat for Research, Development and Innovation (which belongs to MINECO) and by the European Social Fund (ESF) under an FPI fellowship (BES-2015-073768) associated with project FFI2014-51978-C2-1-R. We gratefully acknowledge NVIDIA Corporation for the donation of a GTX Titan X GPU used for this research., Xunta de Galicia; ED431D R2016/046, Xunta e Galicia; ED431B-2017/01
RUC. Repositorio da Universidade da Coruña
Publication Artículo científico (JournalArticle)

How important is syntactic parsing accuracy? An empirical evaluation on rule-based sentiment analysis

  • Gómez-Rodríguez, Carlos
  • Alonso-Alonso, Iago
  • Vilares, David
This version of the article has been accepted for publication, after peer review and is subject to Springer Nature’s AM terms of use, but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: https://doi.org/10.1007/s10462-017-9584-0, [Abstract]: Syntactic parsing, the process of obtaining the internal structure of sentences in natural languages, is a crucial task for artificial intelligence applications that need to extract meaning from natural language text or speech. Sentiment analysis is one example of application for which parsing has recently proven useful. In recent years, there have been significant advances in the accuracy of parsing algorithms. In this article, we perform an empirical, task-oriented evaluation to determine how parsing accuracy influences the performance of a state-of-the-art rule-based sentiment analysis system that determines the polarity of sentences from their parse trees. In particular, we evaluate the system using four well-known dependency parsers, including both current models with state-of-the-art accuracy and more innacurate models which, however, require less computational resources. The experiments show that all of the parsers produce similarly good results in the sentiment analysis task, without their accuracy having any relevant influence on the results. Since parsing is currently a task with a relatively high computational cost that varies strongly between algorithms, this suggests that sentiment analysis researchers and users should prioritize speed over accuracy when choosing a parser; and parsing researchers should investigate models that improve speed further, even at some cost to accuracy., Carlos Gómez-Rodríguez has received funding from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, Grant Agreement No 714150), Ministerio de Economía y Competitividad (FFI2014-51978-C2-2-R), and the Oportunius Program (Xunta de Galicia). Iago Alonso-Alonso was funded by an Oportunius Program Grant (Xunta de Galicia). David Vilares has received funding from the Ministerio de Educación, Cultura y Deporte (FPU13/01180) and Ministerio de Economía y Competitividad (FFI2014-51978-C2-2-R).
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

HEAD-QA: A Healthcare Dataset for Complex Reasoning

  • Vilares, David
  • Gómez-Rodríguez, Carlos
Took place in Florence (Italy) at the 'Fortezza da Basso' from July 28th to August 2nd, 2019., [Absctract]: We present HEAD-QA, a multi-choice question answering testbed to encourage research on complex reasoning. The questions come from exams to access a specialized position in the Spanish healthcare system, and are challenging even for highly specialized humans. We then consider monolingual (Spanish) and cross-lingual (to English) experiments with information retrieval and neural techniques. We show that: (i) HEAD-QA challenges current methods, and (ii) the results lag well behind human performance, demonstrating its usefulness as a benchmark for future work., This work has received support from the TELEPARES-UDC project (FFI2014-51978-C2- 2-R) and the ANSWER-ASAP project (TIN2017- 85160-C2-1-R) from MINECO, from Xunta de Galicia (ED431B 2017/01), and from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, grant agreement No 714150). We thank Mark Anderson for his help with translation fluency evaluation., Xunta de Galicia; ED431B 2017/01
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

Better, Faster, Stronger Sequence Tagging Constituent Parsers

  • Vilares, David
  • Abdou, Mostafa
  • Søgaard, Anders
NAACL-HLT 2019 was held at the Hyatt Regency in Minneapolis from June 2nd through 7th, 2019., [Absctract]: Sequence tagging models for constituent parsing are faster, but less accurate than other types of parsers. In this work, we address the following weaknesses of such constituent parsers: (a) high error rates around closing brackets of long constituents, (b) large label sets, leading to sparsity, and (c) error propagation arising from greedy decoding. To effectively close brackets, we train a model that learns to switch between tagging schemes. To reduce sparsity, we decompose the label set and use multi-task learning to jointly learn to predict sublabels. Finally, we mitigate issues from greedy decoding through auxiliary losses and sentence-level fine-tuning with policy gradient. Combining these techniques, we clearly surpass the performance of sequence tagging constituent parsers on the English and Chinese Penn Treebanks, and reduce their parsing time even further. On the SPMRL datasets, we observe even greater improvements across the board, including a new state of the art on Basque, Hebrew, Polish and Swedish., DV has received support from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, grant agreement No 714150), from the TELEPARES-UDC project (FFI2014-51978-C2-2-R) and the ANSWERASAP project (TIN2017-85160-C2-1-R) from MINECO, and from Xunta de Galicia (ED431B 2017/01). MA and AS are funded by a Google Focused Research Award., Xunta de Galicia; ED431B 2017/01
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

Constituent Parsing as Sequence Labeling

  • Gómez-Rodríguez, Carlos
  • Vilares, David
EMNLP 2018, Square Meeting Center, Brussels. From October 31st through November 4th., [Absctract]: We introduce a method to reduce constituent parsing to sequence labeling. For each word wt, it generates a label that encodes: (1) the number of ancestors in the tree that the words wt and wt+1 have in common, and (2) the nonterminal symbol at the lowest common ancestor. We first prove that the proposed encoding function is injective for any tree without unary branches. In practice, the approach is made extensible to all constituency trees by collapsing unary branches. We then use the PTB and CTB treebanks as testbeds and propose a set of fast baselines. We achieve 90% F-score on the PTB test set, outperforming the Vinyals et al. (2015) sequence-to-sequence parser. In addition, sacrificing some accuracy, our approach achieves the fastest constituent parsing speeds reported to date on PTB by a wide margin., This work has received funding from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, grant agreement No 714150), from the TELEPARESUDC project (FFI2014-51978-C2-2-R) and the ANSWER-ASAP project (TIN2017-85160-C2-1-R) from MINECO, and from Xunta de Galicia (ED431B 2017/01). We gratefully acknowledge NVIDIA Corporation for the donation of a GTX Titan X GPU., Xunta de Galicia; ED431B 2017/01