Pasar al contenido principal

AVANCES EN NUEVOS SISTEMAS DE EXTRACCION DE RESPUESTAS CON ANALISIS SEMANTICO Y APRENDIZAJE PROFUNDO

TIN2017-85160-C2-1-R

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Convocatoria Proyectos I+D+i Retos de Investigación
Año convocatoria 2017
Plan Estatal 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSIDADE DA CORUÑA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 16
Encontrada(s) 1 página(s)

RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

On the Use of Parsing for Named Entity Recognition

  • Alonso, Miguel A.
  • Gómez-Rodríguez, Carlos
  • Vilares, Jesús
[Abstract] Parsing is a core natural language processing technique that can be used to obtain the structure underlying sentences in human languages. Named entity recognition (NER) is the task of identifying the entities that appear in a text. NER is a challenging natural language processing task that is essential to extract knowledge from texts in multiple domains, ranging from financial to medical. It is intuitive that the structure of a text can be helpful to determine whether or not a certain portion of it is an entity and if so, to establish its concrete limits. However, parsing has been a relatively little-used technique in NER systems, since most of them have chosen to consider shallow approaches to deal with text. In this work, we study the characteristics of NER, a task that is far from being solved despite its long history; we analyze the latest advances in parsing that make its use advisable in NER settings; we review the different approaches to NER that make use of syntactic information; and we propose a new way of using parsing in NER based on casting parsing itself as a sequence labeling task., Xunta de Galicia; ED431C 2020/11, Xunta de Galicia; ED431G 2019/01, This work has been funded by MINECO, AEI and FEDER of UE through the ANSWER-ASAP project (TIN2017-85160-C2-1-R); and by Xunta de Galicia through a Competitive Reference Group grant (ED431C 2020/11). CITIC, as Research Center of the Galician University System, is funded by the Consellería de Educación, Universidade e Formación Profesional of the Xunta de Galicia through the European Regional Development Fund (ERDF/FEDER) with 80%, the Galicia ERDF 2014-20 Operational Programme, and the remaining 20% from the Secretaría Xeral de Universidades (Ref. ED431G 2019/01). Carlos Gómez-Rodríguez has also received funding from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, Grant No. 714150).
RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

Sentiment Analysis for Fake News Detection

  • Alonso, Miguel A.
  • Vilares, David
  • Gómez-Rodríguez, Carlos
  • Vilares, Jesús
[Abstract] In recent years, we have witnessed a rise in fake news, i.e., provably false pieces of information created with the intention of deception. The dissemination of this type of news poses a serious threat to cohesion and social well-being, since it fosters political polarization and the distrust of people with respect to their leaders. The huge amount of news that is disseminated through social media makes manual verification unfeasible, which has promoted the design and implementation of automatic systems for fake news detection. The creators of fake news use various stylistic tricks to promote the success of their creations, with one of them being to excite the sentiments of the recipients. This has led to sentiment analysis, the part of text analytics in charge of determining the polarity and strength of sentiments expressed in a text, to be used in fake news detection approaches, either as a basis of the system or as a complementary element. In this article, we study the different uses of sentiment analysis in the detection of fake news, with a discussion of the most relevant elements and shortcomings, and the requirements that should be met in the near future, such as multilingualism, explainability, mitigation of biases, or treatment of multimedia elements., Xunta de Galicia; ED431G 2019/01, Xunta de Galicia; ED431C 2020/11, This work has been funded by FEDER/Ministerio de Ciencia, Innovación y Universidades — Agencia Estatal de Investigación through the ANSWERASAP project (TIN2017-85160-C2-1-R); and by Xunta de Galicia through a Competitive Reference Group grant (ED431C 2020/11). CITIC, as Research Center of the Galician University System, is funded by the Consellería de Educación, Universidade e Formación Profesional of the Xunta de Galicia through the European Regional Development Fund (ERDF/FEDER) with 80%, the Galicia ERDF 2014-20 Operational Programme, and the remaining 20% from the Secretaría Xeral de Universidades (ref. ED431G 2019/01). David Vilares is also supported by a 2020 Leonardo Grant for Researchers and Cultural Creators from the BBVA Foundation. Carlos Gómez-Rodríguez has also received funding from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, grant No. 714150)
RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

Isolation Number versus Domination Number of Trees

  • Lemańska, Magdalena
  • Souto Salorio, María José
  • Dapena, Adriana
  • Vázquez Araújo, Francisco Javier
[Abstract] If 𝐺 = (Vɢ,Eɢ) is a graph of order n, we call 𝑆 ⊆ Vɢ an isolating set if the graph induced by Vɢ − Nɢ[𝑆] contains no edges. The minimum cardinality of an isolating set of 𝐺 is called the isolation number of 𝐺, and it is denoted by 𝜄(𝐺). It is known that 𝜄(𝐺) ≤ ⁿ⁄₃ and the bound is sharp. A subset 𝑆 ⊆ Vɢ is called dominating in 𝐺 if Nɢ[𝑆] = Vɢ. The minimum cardinality of a dominating set of 𝐺 is the domination number, and it is denoted by 𝛾(𝐺). In this paper, we analyze a family of trees 𝑇 where 𝜄(𝑇) = 𝛾(𝑇), and we prove that 𝜄(T) = ⁿ⁄₃ implies 𝜄(𝑇) = 𝛾(𝑇). Moreover, we give different equivalent characterizations of such graphs and we propose simple algorithms to build these trees from the connections of stars., CITIC, as Research Center accredited by Galician University System, is funded by "Consellería de Cultura, Educación e Universidade from Xunta de Galicia", supported in an 80% through ERDF Funds, ERDF Operational Programme Galicia 2014-2020, and the remaining 20% by "Secretaría Xeral de Universidades (Grant ED431G 2019/01). This research was also funded by Agencia Estatal de Investigación of Spain (PID2019-104958RB-C42 and TIN2017-85160-C2-1-R) and ERDF funds of the EU (AEI/FEDER, UE)., Xunta de Galicia; ED431G 2019/01
RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

Bertinho: Galician BERT Representations, Bertinho: Representaciones BERT para el gallego

  • Vilares, David
  • García, Marcos
  • Gómez-Rodríguez, Carlos
[Abstract]: 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., [Resumen]: Este artículo presenta un modelo BERT monolingüe para el gallego. Nos basamos en la tendencia actual que ha demostrado que es posible crear modelos BERT monolingües robustos incluso para aquellos idiomas para los que hay una relativa escasez de recursos, funcionando éstos mejor que el modelo BERT multilingüe oficial (mBERT). Concretamente, liberamos dos modelos monolingües para el gallego, creados con 6 y 12 capas de transformers, respectivamente, y entrenados con una limitada cantidad de recursos (~45 millones de palabras sobre una única GPU de 24GB.) Para evaluarlos realizamos un conjunto exhaustivo de experimentos en tareas como análisis morfosintáctico, análisis sintáctico de dependencias o reconocimiento de entidades. Para ello, abordamos estas tareas como etiquetado de secuencias, con el objetivo de ejecutar los modelos BERT sin la necesidad de incluir ninguna capa adicional (únicamente se a~nade la capa de salida encargada de transformar las representaciones contextualizadas en la etiqueta predicha). Los experimentos muestran que nuestros modelos, especialmente el de 12 capas, mejoran los resultados de mBERT en la mayor parte de las tareas., This work has received funding from the European Research Council (ERC), which has funded this research under the European Union's Horizon 2020 research and innovation programme (FASTPARSE, grant agreement No 714150), from MINECO (ANSWER-ASAP, TIN2017-85160-C2-1-R), from Xunta de Galicia (ED431C 2020/11), from Centro de Investigación de Galicia `CITIC', funded by Xunta de Galicia and the European Union (European Regional Development Fund- Galicia 2014-2020 Program), by grant ED431G 2019/01, and by Centro Singular de Investigación en Tecnoloxías Intelixentes (CiTIUS), ERDF 2014-2020: Call ED431G 2019/04. DV is supported by a 2020 Leonardo Grant for Researchers and Cultural Creators from the BBVA Foundation. MG is supported by a Ramón y Cajal grant (RYC2019-028473-I)., Xunta de Galicia; ED431C 2020/11, Xunta de Galicia; ED431G 2019/01, Xunta de Galicia; D431G 2019/04
RUC. Repositorio da Universidade da Coruña
Publicación 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
Publicación Artículo científico (JournalArticle)

Towards Robust Word Embeddings for Noisy Texts

  • Doval, Yerai
  • Vilares, Jesús
  • Gómez-Rodríguez, Carlos
[Abstract] 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 these types of noisy texts at the word embedding level that goes beyond the support for out-of-vocabulary words., Yerai Doval has been supported by the Spanish Ministry of Economy, Industry and Competitiveness (MINECO) through the ANSWER-ASAP project (TIN2017-85160-C2-2-R); by the Spanish State Secretariat for Research, Development and Innovation (which belongs to MINECO) and the European Social Fund (ESF) through a FPI fellowship (BES-2015-073768) associated with TELEPARES project (FFI2014-51978-C2-1-R); and by the Xunta de Galicia through TELGALICIA research network (ED431D 2017/12). The work of Jesús Vilares and Carlos Gómez-Rodríguez has also been funded by MINECO through the ANSWER-ASAP project (TIN2017-85160-C2-1-R in this case); and by Xunta de Galicia through a Group with Potential for Growth grant (ED431B 2017/01), a Competitive Reference Group grant (ED431C 2020/11), and a Remarkable Research Centre grant for the CITIC research centre (ED431G/01), the latter co-funded by EU with ERDF funding. Finally, Carlos Gómez-Rodríguez has also received funding from the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, Grant No. 714150), Xunta de Galicia; ED431D 2017/12, Xunta de Galicia; ED431B 2017/01, Xunta de Galicia; ED431C 2020/11, Xunta de Galicia; ED431G/01
RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

Optimality of syntactic dependency distances

  • Ferrer-i-Cancho, Ramon
  • Gómez-Rodríguez, Carlos
  • Esteban, Juan Luis
  • Alemany-Puig, Lluís
[Abstract]: 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 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: 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 hierarchical ranking of languages by their degree of optimization. The 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., We are grateful to C. Bentz, N. Català, M. H. Christiansen, M. Gustison, and A. Hernández-Fernández for helpful comments. R.F.C. and L.A.P. are supported by Grant No. TIN2017-89244-R from MINECO (Ministerio de Economía, Industria y Competitividad). R.F.C. is also supported by the recognition No. 2017SGR-856 (MACDA) from AGAUR (Generalitat de Catalunya). L.A.P. is also supported by Secretaria d’Universitats i Recerca de la Generalitat de Catalunya and the Social European Fund. C.G.R. is supported by the European Research Council (ERC), under the European Union’s Horizon 2020 research and innovation program (FASTPARSE, Grant No. 714150), the ANSWER-ASAP project (No. TIN2017-85160-C2-1-R) and SCANNER-UDC project (No. PID2020-113230RB-C21) from ERDF/MICINN-AEI, and Xunta de Galicia (ED431C 2020/11 and an Oportunius program grant to complement ERC grants); the CITIC research center is funded by ERDF and Xunta de Galicia (ERDFGalicia 2014-2020 program, Grant No. ED431G 2019/01). J.L.E. is funded by Grants No. TIN2016-76573-C2-1-P and No. PID2019-109137GB-C22 from MINECO. R.F.C. designed the research; R.F.C and C.G.R. wrote the paper; C.G.R. collected and preprocessed data; C.G.R. and R.F.C. analyzed the data; R.F.C, C.G.R., and J.L.E developed the mathematical arguments; C.G.R., J.L.E., and L.A.P. wrote the core code; all authors verified the mathematical arguments., Xunta de Galicia; ED431C 2020/11, Xunta de Galicia; ED431G 2019/01, Agència de Gestió d'Ajuts Universitaris i de Recerca; 2017SGR-856
RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

Classification of the relative positions between a small ellipsoid and an elliptic paraboloid

  • Brozos-Vázquez, Miguel
  • Pereira-Sáez, María José
  • Souto Salorio, María José
  • Tarrío-Tobar, Ana D.
©2019. 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 Brozos-Vázquez, M., Pereira-Sáez, M. J., Souto-Salorio, M. J., & Tarrío-Tobar, A. D. (2019). “Classification of the relative positions between a small ellipsoid and an elliptic paraboloid” has been accepted for publication in Computer Aided Geometric Design, 72, 34–48. The Version of Record is available online at https://doi.org/10.1016/j.cagd.2019.05.002., [Abstract]: We classify all the relative positions between an ellipsoid and an elliptic paraboloid when the ellipsoid is small in comparison with the paraboloid (small meaning that the two surfaces cannot be tangent at two points simultaneously when one is moved with respect to the other). This provides an easy way to detect contact between the two surfaces by a direct analysis of the coefficients of a fourth degree polynomial., The authors wish to thank the referees for extremely valuable comments and suggestions, which were essential to improve the final version of the paper. Supported by Projects ED431F 2017/03, TIN2017-85160-C2-1-R, MTM2016-75897-P and MTM2016-78647-P (AEI/FEDER, UE)., Xunta de Galicia; ED431F 2017/03
RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

Discovering Topics in Twitter About the COVID-19 Outbreak in Spain, Descubriendo temas en Twitter sobre el brote del COVID-19 en España

  • Agüero-Torales, Marvin M.
  • Vilares, David
  • López-Herrera, Antonio G.
[Resumen] En este trabajo, analizamos lo que los usuarios han estado discutiendo en Twitter durante el comienzo de la pandemia causada por el COVID-19. Concretamente, analizamos tres fases diferenciadas de la crisis del COVID-19 en España: el propio tiempo de pre-crisis, el estallido de la enfermedad y el confinamiento. Para llevar esto a cabo, primero recolectamos una gran cantidad de tuits que son preprocesados. A continuación, agrupamos los tuits en distintas temáticas usando un modelo de Latent Dirichlet Allocation, y definimos estrategias generativas y discriminativas para extraer las palabras clave y oraciones más representativas para cada tema. Finalmente, incluimos un exhaustivo análisis cualitativo sobre dichos temas, y cómo estos se corresponden con distintas problemáticas surgidas en España en distintos momentos de la crisis., [Abstract] In this work, we apply topic modeling to study what users have been discussing in Twitter during the beginning of the COVID-19 pandemic. More particularly, we explore the period of time that includes three differentiated phases of the COVID-19 crisis in Spain: the pre-crisis time, the outbreak, and the beginning of the lockdown. To do so, we first collect a large corpus of Spanish tweets and clean them. Then, we cluster the tweets into topics using a Latent Dirichlet Allocation model, and define generative and discriminative routes to later extract the most relevant keywords and sentences for each topic. Finally, we provide an exhaustive qualitative analysis about how such topics correspond to the situation in Spain at different stages of the crisis., MMAT has been partially funded by Barcelona Supercomputing Center (BSC) through the Spanish Plan for advancement of Language Technologies `Plan TL' and the Secretaría de Estado de Digitalización e Inteligencia Artificial (SEDIA). DV is supported by MINECO (TIN2017-85160-C2-1-R), by Xunta de Galicia (ED431C 2020/11), by Centro de Investigación de Galicia `CITIC' (European Regional Development Fund-Galicia 2014-2020 Program, ED431G 2019/01), and by a 2020 Leonardo Grant for Researchers and Cultural Creators from the BBVA Foundation, Xunta de Galicia; ED431C 2020/11, Xunta de Galicia; ED431G 2019/01
RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

Multitask Pointer Network for Multi-Representational Parsing

  • Fernández-González, Daniel
  • Gómez-Rodríguez, Carlos
Financiado para publicación en acceso aberto: Universidade da Coruña/CISUG, [Abstract] Dependency and constituent trees are widely used by many artificial intelligence applications for representing the syntactic structure of human languages. Typically, these structures are separately produced by either dependency or constituent parsers. In this article, 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., We acknowledge the European Research Council (ERC), which has funded this research under the European Union’s Horizon 2020 research and innovation programme (FASTPARSE, grant agreement No 714150), ERDF/MICINN-AEI (ANSWER-ASAP, TIN2017-85160-C2-1-R; SCANNER-UDC, PID2020-113230RB-C21), Xunta de Galicia, Spain (ED431C 2020/11), and Centro de Investigación de Galicia “CITIC”, funded by Xunta de Galicia, Spain and the European Union (ERDF - Galicia 2014–2020 Program), by grant ED431G 2019/01. Funding for open access charge: Universidade da Coruña / CISUG, Xunta de Galicia; ED431C 2020/11, Xunta de Galicia; ED431G 2019/01
RUC. Repositorio da Universidade da Coruña
Publicación Artículo científico (JournalArticle)

Faster shift-reduce constituent parsing with a non-binary, bottom-up strategy

  • Fernández-González, Daniel
  • Gómez-Rodríguez, Carlos
© 2019. This is the final peer-reviewed manuscript that was accepted for publication at Artificial Intelligence and made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/ licenses/by-nc-nd/4.0/. This may not reflect subsequent changes resulting from the publishing process such as editing, formatting, pagination, and other quality control mechanisms. The final journal publication is available at https://doi.org/10.1016/j.artint.2019.07.006, [Absctract]: An increasingly wide range of artificial intelligence applications rely on syntactic information to process and extract meaning from natural language text or speech, with constituent trees being one of the most widely used syntactic formalisms. To produce these phrase-structure representations from sentences in natural language, shift-reduce constituent parsers have become one of the most efficient approaches. Increasing their accuracy and speed is still one of the main objectives pursued by the research community so that artificial intelligence applications that make use of parsing outputs, such as machine translation or voice assistant services, can improve their performance. With this goal in mind, we propose in this article a novel non-binary shift-reduce algorithm for constituent parsing. Our parser follows a classical bottom-up strategy but, unlike others, it straightforwardly creates non-binary branchings with just one transition, instead of requiring prior binarization or a sequence of binary transitions, allowing its direct application to any language without the need of further resources such as percolation tables. As a result, it uses fewer transitions per sentence than existing transition-based constituent parsers, becoming the fastest such system and, as a consequence, speeding up downstream applications. Using static oracle training and greedy search, the accuracy of this novel approach is on par with state-of-the-art transition-based constituent parsers and outperforms all top-down and bottom-up greedy shift-reduce systems on the Wall Street Journal section from the English Penn Treebank and the Penn Chinese Treebank. Additionally, we develop a dynamic oracle for training the proposed transition-based algorithm, achieving further improvements in both benchmarks and obtaining the best accuracy to date on the Penn Chinese Treebank among greedy shift-reduce parsers., 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 ANSWER-ASAP project (TIN2017-85160-C2-1-R) from MINECO, and from Xunta de Galicia (ED431B 2017/01)., Xunta de Galicia; ED431B 2017/01
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación Artículo científico (JournalArticle)

Faster shift-reduce constituent parsing with a non-binary, bottom-up strategy

  • Fernández González, Daniel
  • Gómez Rodríguez, Carlos
An increasingly wide range of artificial intelligence applications rely on syntactic information to process and extract meaning from natural language text or speech, with constituent trees being one of the most widely used syntactic formalisms. To produce these phrase-structure representations from sentences in natural language, shift-reduce constituent parsers have become one of the most efficient approaches. Increasing their accuracy and speed is still one of the main objectives pursued by the research community so that artificial intelligence applications that make use of parsing outputs, such as machine translation or voice assistant services, can improve their performance. With this goal in mind, we propose in this article a novel non-binary shift-reduce algorithm for constituent parsing. Our parser follows a classical bottom-up strategy but, unlike others, it straightforwardly creates non-binary branchings with just one transition, instead of requiring prior binarization or a sequence of binary transitions, allowing its direct application to any language without the need of further resources such as percolation tables. As a result, it uses fewer transitions per sentence than existing transition-based constituent parsers, becoming the fastest such system and, as a consequence, speeding up downstream applications. Using static oracle training and greedy search, the accuracy of this novel approach is on par with state-of-the-art transition-based constituent parsers and outperforms all top-down and bottom-up greedy shift-reduce systems on the Wall Street Journal section from the English Penn Treebank and the Penn Chinese Treebank. Additionally, we develop a dynamic oracle for training the proposed transition-based algorithm, achieving further improvements in both benchmarks and obtaining the best accuracy to date on the Penn Chinese Treebank among greedy shift-reduce parsers., Agencia Estatal de Investigación | Ref. TIN2017-85160-C2-1-R, Xunta de Galicia | Ref. ED431B 2017/01
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación 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
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación Artículo científico (JournalArticle)

Adaptive scheduling for adaptive sampling in pos taggers construction

  • Vilares Ferro, Manuel
  • Darriba Bilbao, Víctor Manuel
  • Vilares Ferro, Jesús
We introduce an adaptive scheduling for adaptive sampling as a novel way of machine learning in the construction of part-of-speech taggers. The goal is to speed up the training on large data sets, without significant loss of performance with regard to an optimal configuration. In contrast to previous methods using a random, fixed or regularly rising spacing between the instances, ours analyzes the shape of the learning curve geometrically in conjunction with a functional model to increase or decrease it at any time. The algorithm proves to be formally correct regarding our working hypotheses. Namely, given a case, the following one is the nearest ensuring a net gain of learning ability from the former, it being possible to modulate the level of requirement for this condition. We also improve the robustness of sampling by paying greater attention to those regions of the training data base subject to a temporary inflation in performance, thus preventing the learning from stopping prematurely. The proposal has been evaluated on the basis of its reliability to identify the convergence of models, corroborating our expectations. While a concrete halting condition is used for testing, users can choose any condition whatsoever to suit their own specific needs., Agencia Estatal de Investigación | Ref. TIN2017-85160-C2-1-R, Agencia Estatal de Investigación | Ref. TIN2017-85160-C2-2-R, Xunta de Galicia | Ref. ED431C 2018/50, Xunta de Galicia | Ref. ED431D 2017/12
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación Artículo científico (JournalArticle)

Absolute convergence and error thresholds in non-active adaptive sampling

  • Vilares Ferro, Manuel
  • Darriba Bilbao, Víctor Manuel
  • Vilares Ferro, Jesús
Financiado para publicación en acceso aberto: Universidade de Vigo/CISUG, Non-active adaptive sampling is a way of building machine learning models from a training data base which are supposed to dynamically and automatically derive guaranteed sample size. In this context and regardless of the strategy used in both scheduling and generating of weak predictors, a proposal for calculating absolute convergence and error thresholds is described. We not only make it possible to establish when the quality of the model no longer increases, but also supplies a proximity condition to estimate in absolute terms how close it is to achieving such a goal, thus supporting decision making for fine-tuning learning parameters in model selection. The technique proves its correctness and completeness with respect to our working hypotheses, in addition to strengthening the robustness of the sampling scheme. Tests meet our expectations and illustrate the proposal in the domain of natural language processing, taking the generation of part-of-speech taggers as case study., Agencia Estatal de Investigación | Ref. TIN2017-85160-C2-1-R, Agencia Estatal de Investigación | Ref. TIN2017-85160-C2-2-R, Agencia Estatal de Investigación | PID2020-113230RB-C21, Agencia Estatal de Investigación | PID2020-113230RB-C22, Xunta de Galicia | Ref. ED431C 2018/50
Investigo. Repositorio Institucional de la Universidade de Vigo
Publicación Artículo científico (JournalArticle)

Multitask pointer network for multi-representational parsing

  • Fernández González, Daniel
  • Gómez Rodríguez, Carlos
Dependency and constituent trees are widely used by many artificial intelligence applications for representing the syntactic structure of human languages. Typically, these structures are separately produced by either dependency or constituent parsers. In this article, 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., Agencia Estatal de Investigación | Ref. TIN2017-85160-C2-1-R, Agencia Estatal de Investigación | Ref. PID2020-113230RB-C21, Xunta de Galicia | Ref. ED431C 2020/11, Xunta de Galicia | Ref. ED431G 2019/01, Universidade da Coruña/CISUG