Skip to main content

Resultados relacionados

Resultados totales (Incluyendo duplicados): 14
Encontrada(s) 1 página(s)

RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

EN-ES-CS: An English-Spanish Code-Switching Twitter Corpus for Multilingual Sentiment Analysis

  • Vilares, David
  • Alonso, Miguel A.
  • Gómez-Rodríguez, Carlos
[Abstract]: Code-switching texts are those that contain terms in two or more different languages, and they appear increasingly often in social media. The aim of this paper is to provide a resource to the research community to evaluate the performance of sentiment classification techniques on this complex multilingual environment, proposing an English-Spanish corpus of tweets with code-switching (EN-ES-CS CORPUS). The tweets are labeled according to two well-known criteria used for this purpose: SentiStrength and a trinary scale (positive, neutral and negative categories). Preliminary work on the resource is already done, providing a set of baselines for the research community., This research is supported by the Ministerio de Economía y Competitividad (FFI2014-51978-C2) and Xunta de Galicia (R2014/034). David Vilares is funded by the Ministerio de Educación, Cultura y Deporte (FPU13/01180). Carlos Gómez-Rodríguez is funded by an Oportunius program grant., Xunta de Galicia; R2014/034
RUC. Repositorio da Universidade da Coruña
Publication Artículo científico (JournalArticle)

A linguistic approach for determining the topics of Spanish Twitter messages

  • Vilares, David
  • Alonso, Miguel A.
  • Gómez-Rodríguez, Carlos
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: Vilares, D., Alonso, M. A., & Gómez-Rodríguez, C. (2015). ‘A linguistic approach for determining the topics of Spanish Twitter messages’ has been accepted for publication in Journal of Information Science, 41(2), 127-145. Copyright © 2014 The Authors. DOI: https://doi.org/10.1177/0165551514561652., [Abstract]: The vast number of opinions and reviews provided in Twitter is helpful in order to make interesting findings about a given industry, but given the huge number of messages published every day, it is important to detect the relevant ones. In this respect, the Twitter search functionality is not a practical tool when we want to poll messages dealing with a given set of general topics. This article presents an approach to classify Twitter messages into various topics. We tackle the problem from a linguistic angle, taking into account part-of-speech, syntactic and semantic information, showing how language processing techniques should be adapted to deal with the informal language present in Twitter messages. The TASS 2013 General corpus, a collection of tweets that has been specifically annotated to perform text analytics tasks, is used as the dataset in our evaluation framework. We carry out a wide range of experiments to determine which kinds of linguistic information have the greatest impact on this task and how they should be combined in order to obtain the best-performing system. The results lead us to conclude that relating features by means of contextual information adds complementary knowledge over pure lexical models, making it possible to outperform them on standard metrics for multilabel classification tasks., The research reported in this article was partially funded by Ministerio de Economía y Competitividad and FEDER (grant TIN2010-18552-C03-02), Ministerio de Educación, Cultura y Deporte (FPU13/01180) and by Xunta de Galicia (Grants CN2012/008, CN2012/319)., Xunta de Galicia; CN2012/008, Xunta de Galicia; CN2012/319
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

LyS at TASS 2014: A Prototype for Extracting and Analysing Aspects from Spanish tweets, LyS en TASS 2014: Un prototipo para la extracción y análisis de aspectos en tuits.

  • Vilares, David
  • Doval, Yerai
  • Alonso, Miguel A.
  • Gómez-Rodríguez, Carlos
[Abstract]: This paper describes our participation at the third edition of the work- shop on Sentiment Analysis focused on Spanish tweets, tass 2014. This year’s eval- uation campaign includes four challenges: (1) global sentiment analysis, (2) topic classification, (3) aspect-extraction and (4) aspect-based sentiment analysis. Tasks 1 and 2 are addressed from a machine learning approach, using several linguistic resources and other information extracted from the training corpus to feed to a su- pervised classifier. With respect to task 3, we develop a naive approach, collecting a set of representations to identify the predefined aspects requested by the organisers. Finally, task 4 uses heuristics to identify the scope of each aspect, to then classify their sentiment via a supervised classifier. The experimental results are promising and will serve us as the starting point to develop more complex techniques., [Resumen]: Este artículo describe nuestra participación en la tercera edición del taller de análisis del sentimiento de tuits escritos en castellano, el tass 2014. En la evaluación competitiva de este año, se han propuesto cuatro retos: (1) análisis del sentimiento a nivel global, (2) clasificación de tópicos, (3) extracción de aspectos y (4) análisis del sentimiento a nivel aspectual. Para las tareas 1 y 2 empleamos una aproximación basada en aprendizaje automático, donde distintos recursos lingüísticos e información extraída del conjunto de entrenamiento son utilizados para entrenar un clasificador supervisado. Para abordar la tarea 3, nuestra aproximación recolecta una lista de representaciones que es empleada para identificar los aspectos requeridos por los organizadores. Por último, la tarea 4 delega en heurísticas para identificar el alcance de cada aspecto, para después determinar su sentimiento a través de un clasificador supervisado. Los resultados experimentales son prometedores y nos servirán para desarrollar técnicas más complejas en el futuro., Research reported in this paper has been partially funded by Ministerio de Educación, Cultura y Deporte (FPU13/01180) and Xunta de Galicia (Grant CN2012/008)., Xunta de Galicia; CN2012/008
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

LyS at TASS 2015: Deep Learning Experiments for Sentiment Analysis on Spanish Tweets

  • Vilares, David
  • Doval, Yerai
  • Alonso, Miguel A.
  • Gómez-Rodríguez, Carlos
[Abstract]: This paper describes the participation of the LyS group at tass 2015. In this year’s edition, we used a long short-term memory neural network to address the two proposed challenges: (1) sentiment analysis at a global level and (2) aspect-based sentiment analysis on football and political tweets. The performance of this deep learning approach is compared to our last-year model, based on a square-regularized logistic regression. Experimental results show that strategies such as unsupervised pre-training, sentiment-specific word embedding or modifying the current architec- ture might be needed to achieve state-of-the-art results., [Resumen]: Este art´ıculo describe la participaci´on del grupo LyS en el tass 2015. En la edici´on de este a˜no, hemos utilizado una red neuronal denominada long short- term memory para abordar los dos retos propuestos: (1) an´alisis del sentimiento a nivel global y (2) an´alisis del sentimiento a nivel de aspectos sobre tuits futbol´ısticos y de pol´ıtica. El rendimiento obtenido por esta red de aprendizaje profundo es comparado con el de nuestro sistema del a˜no pasado, una regresi´on log´ıstica con una regularizaci´on cuadr´atica. Los resultados experimentales muestran que es necesario incluir estrategias como pre-entrenamiento no supervisado, t´ecnicas espec´ıficas para representar palabras como vectores o modificar la arquitectura actual para alcanzar resultados acordes con el estado del arte., This research is supported by the Ministerio de Economía y Competitividad y FEDER (FFI2014-51978-C2) and Xunta de Galicia (R2014/034). The first author is funded by the Ministerio de Educación, Cultura y Deporte (FPU13/01180)., Xunta de Galicia; R2014/034
RUC. Repositorio da Universidade da Coruña
Publication Artículo científico (JournalArticle)

The megaphone of the people? Spanish SentiStrength for real-time analysis of political tweets

  • Vilares, David
  • Thelwall, Mike
  • Alonso, Miguel A.
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: Vilares, D., Thelwall, M., & Alonso, M. A. (2015). ‘The megaphone of the people? Spanish SentiStrength for real-time analysis of political tweets’ has been accepted for publication in Journal of Information Science, 41(6), 799-813. Copyright © 2015 The Authors. DOI: https://doi.org/10.1177/0165551515598926., [Abstract]: Twitter is an important platform for sharing opinions about politicians, parties and political decisions. These opinions can be exploited as a source of information to monitor the impact of politics on society. This article analyses the sentiment of 2,704,523 tweets referring to Spanish politicians and parties from a month in 2014–2015. The article makes three specific contributions: (a) enriching SentiStrength, a fast unsupervised sentiment strength detection system, for Spanish political tweeting; (b) analysing how linguistic phenomena such as negation, idioms and character duplication influence Spanish sentiment strength detection accuracy; and (c) analysing Spanish political tweets to rank political leaders, parties and personalities for popularity. Sentiment in Twitter for key politicians broadly reflects the main official polls for popularity but not for voting intention. In addition, the data suggests that the primary role of Twitter in politics is to select and amplify political events published by traditional media., This research is partially supported by the Ministerio de Economía y Competitividad (FFI2014-51978-C2) and Xunta de Galicia (R2014/034). The first author is partially funded by the Ministerio de Educación, Cultura y Deporte (FPU13/01180) and Inditex-UDC 2014 grants for international visits. The authors thank the annotators for help with this research., Xunta de Galicia; R2014/034
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

Identificación Automática del Idioma en Twitter: Adaptación de Identificadores del Estado del Arte al Contexto Ibérico, Automatic Language Identification in Twitter: Adapting State-of-the-Art Identifiers to the Iberian Context

  • Doval, Yerai
  • Vilares, David
  • Vilares, Jesús
[Abstract]: We describe here our partipation in TweetLID. After having studied the problem of language identification, the resources available, and designed a text conflation approach for this kind of tasks, we joined the competition with two sys- tems: the first one was based in the guesser langdetect, re-trained and adapted in order to work with conflated text; the second one was an approach based on majo- rity vote which used a set of re-trained and adapted classifiers. Results are analized both globally and at language and tweet-type levels., [Resumen]: Describimos aquí nuestra participación en el TweetLID. Tras estudiar la problemática de la identificación del idioma, los recursos disponibles y diseñar una solución para la normalización del texto en este tipo de tareas, presentamos dos sis- temas a competición: el primero basado en el clasificador langdetect, re-entrenado y adaptado para trabajar con texto normalizado; el segundo, una solución basada en la votación de clasificadores individuales igualmente re-entrenados y adaptados. Los resultados son analizados tanto globalmente como por idioma y tipo de tuit., Trabajo parcialmente financiado por el Ministeriode Economía y Competitividad y FEDER (TIN2010-18552-C03-02), Ministerio de Educación, Cultura y Deporte (Beca FPU ref. 13/01180) y Xunta de Galicia (CN2012/008)., Xunta de Galicia; CN2012/008
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)

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 non-projective greedy dependency parser with bidirectional LSTMs

  • Vilares, David
  • Gómez-Rodríguez, Carlos
[Abstract]: The LyS-FASTPARSE team present BIST-COVINGTON, a neural implementation of the Covington (2001) algorithm for non-projective dependency parsing. The bidirectional LSTM approach by Kiperwasser and Goldberg (2016) is used to train a greedy parser with a dynamic oracle to mitigate error propagation. The model participated in the CoNLL 2017 UD Shared Task. In spite of not using any ensemble methods and using the baseline segmentation and PoS tagging, the parser obtained good results on both macro-average LAS and UAS in the big treebanks category (55 languages), ranking 7th out of 33 teams. In the all treebanks category (LAS and UAS) we ranked 16th and 12th. The gap between the all and big categories is mainly due to the poor performance on four parallel PUD treebanks, suggesting that some ‘suffixed’ treebanks (e.g. Spanish-AnCora) perform poorly on cross-treebank settings, which does not occur with the corresponding ‘unsuffixed’ treebank (e.g. Spanish). By changing that, we obtain the 11th best LAS among all runs (official and unofficial). The code is made available at https://github.com/CoNLL-UD-2017/LyS-FASTPARSE., David Vilares is funded by an FPU Grant 13/01180. 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). Both authors have received funding from the TELEPARES-UDC project from MINECO.
RUC. Repositorio da Universidade da Coruña
Publication Artículo científico (JournalArticle)

Universal, unsupervised (rule-based), uncovered sentiment analysis

  • Vilares, David
  • Gómez-Rodríguez, Carlos
  • Alonso, Miguel A.
© 2017. 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: Vilares, D., Gómez-Rodríguez, C. and Alonso, M.A. (2017) ‘Universal, unsupervised (rule-based), uncovered sentiment analysis’ has been accepted for publication in Knowledge-Based Systems, 118, pp. 45–55. The Version of Record is available online at https://doi.org/10.1016/j.knosys.2016.11.014., [Abstract]: We present a novel unsupervised approach for multilingual sentiment analysis driven by compositional syntax-based rules. On the one hand, we exploit some of the main advantages of unsupervised algorithms: (1) the interpretability of their output, in contrast with most supervised models, which behave as a black box and (2) their robustness across different corpora and domains. On the other hand, by introducing the concept of compositional operations and exploiting syntactic information in the form of universal dependencies, we tackle one of their main drawbacks: their rigidity on data that are structured differently depending on the language concerned. Experiments show an improvement both over existing unsupervised methods, and over state-of-the-art supervised models when evaluating outside their corpus of origin. Experiments also show how the same compositional operations can be shared across languages. The system is available at http://www.grupolys.org/software/UUUSA/, This research is supported by the Ministerio de Economía y Competitividad (FFI2014-51978-C2). David Vilares is funded by the Ministerio de Educación, Cultura y Deporte (FPU13/01180). Carlos Gómez-Rodríguez is funded by an Oportunius program grant (Xunta de Galicia). We thank Roman Klinger for his help in obtaining the German data.
RUC. Repositorio da Universidade da Coruña
Publication Contribución de congreso (ConferenceOutput)

Sentiment Analysis on Monolingual, Multilingual and Code-Switching Twitter Corpora

  • Vilares, David
  • Alonso, Miguel A.
  • Gómez-Rodríguez, Carlos
[Abstract]: We address the problem of performing po- larity classification on Twitter over differ- ent languages, focusing on English and Spanish, comparing three techniques: (1) a monolingual model which knows the language in which the opinion is written, (2) a monolingual model that acts based on the decision provided by a language iden- tification tool and (3) a multilingual model trained on a multilingual dataset that does not need any language recognition step. Results show that multilingual models are even able to outperform the monolingual models on some monolingual sets. We introduce the first code-switching corpus with sentiment labels, showing the robust- ness of a multilingual approach., This research is supported by the Ministerio de Economía y Competitividad (FFI2014-51978-C2) and Xunta de Galicia (R2014/034). The first authoris funded by the Ministerio de Educación, Cultura y Deporte (FPU13/01180)., Xunta de Galicia; R2014/034
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)

LyS at SemEval-2016 Task 4: Exploiting Neural Activation Values for Twitter Sentiment Classification and Quantification

  • Vilares, David
  • Doval, Yerai
  • Alonso, Miguel A.
  • Gómez-Rodríguez, Carlos
[Abstract]: In this paper we describe our deep learning approach for solving both two-, three- and fiveclass tweet polarity classification, and twoand five-class quantification. We first trained a convolutional neural network using pretrained Twitter word embeddings, so that we could extract the hidden activation values from the hidden layers once some input had been fed to the network. These values were then used as features for a support vector machine in both the classification and quantification subtasks, together with additional linguistic information in the former scenario. The results obtained for the classification subtasks show that this approach performs better than a single convolutional network, and for the quantification part it also yields good results. Official rankings locate us: 2nd (practically tied with 1st) for the binary classification task, 2nd for binary quantification and 4th (practically tied with 3rd) for the five-class polarity classification challenge., This research is supported by the Ministerio de Economía y Competitividad (FFI2014-51978-C2). David Vilares is funded by the Ministerio de Educación, Cultura y Deporte (FPU13/01180). Yerai Doval is funded by the Ministerio de Economía y Competitividad (BES-2015-073768). Carlos Gómez-Rodríguez is funded by an Oportunius program grant (Xunta de Galicia)
RUC. Repositorio da Universidade da Coruña
Publication Artículo científico (JournalArticle)

Supervised sentiment analysis in multilingual environments

  • Vilares, David
  • Alonso, Miguel A.
  • Gómez-Rodríguez, Carlos
© 2017. 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: Vilares, D., Alonso, M.A. and Gómez-Rodríguez, C. (2017) ‘Supervised sentiment analysis in multilingual environments’ has been accepted for publication in Information Processing & Management, 53(3), pp. 595–607. The Version of Record is available online at https://doi.org/10.1016/j.ipm.2017.01.004., [Abstract]: This article tackles the problem of performing multilingual polarity classification on Twitter, comparing three techniques: (1) a multilingual model trained on a multilingual dataset, obtained by fusing existing monolingual resources, that does not need any language recognition step, (2) a dual monolingual model with perfect language detection on monolingual texts and (3) a monolingual model that acts based on the decision provided by a language identification tool. The techniques were evaluated on monolingual, synthetic multilingual and code-switching corpora of English and Spanish tweets. In the latter case we introduce the first code-switching Twitter corpus with sentiment labels. The samples are labelled according to two well-known criteria used for this purpose: the SentiStrength scale and a trinary scale (positive, neutral and negative categories). The experimental results show the robustness of the multilingual approach (1) and also that it outperforms the monolingual models on some monolingual datasets., This research was supported by the Ministerio de Economía y Competitividad (FFI2014-51978-C2) and Xunta de Galicia (R2014/034). David Vilares is funded by the Ministerio de Educación, Cultura y Deporte (FPU13/01180). Carlos Gómez-Rodríguez is funded by an Oportunius program grant (Xunta de Galicia)., Xunta de Galicia; R2014/034