Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 6
Encontrada(s) 1 página(s)

RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Overview of HOPE at IberLEF 2023: Multilingual Hope Speech Detection, Resumen de la tarea HOPE en IberLEF 2023: Detección Multilingüe de Discurso Esperanzador

  • Jiménez Zafra, Salud M.
  • García Cumbreras, Miguel Ángel
  • García-Baena, Daniel
  • García-Díaz, José Antonio
  • Chakravarthi, Bharathi Raja
  • Valencia García, Rafael
  • Ureña López, Luis Alfonso
Hope speech is the speech that is able to relax hostile environments and that helps, inspires and encourages people in times of illness, stress, loneliness or depression. Its automatic recognition can have a very significant effect fighting against sexual and racial discrimination or fostering less belligerent environments. In contrast to identifying and censoring negative or hate speech, hope speech detection is focused on recognizing and promoting positive speech online. In this paper we present an overview of the IberLEF 2023 shared task, HOPE: Multilingual Hope Speech Detection, consisting of identifying whether texts written in English or Spanish contain hope speech or not. The competition was organized through CodaLab and attracted 50 teams that registered. Finally, 12 submitted results and 8 presented working notes describing their systems.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Spanish hate-speech detection in football, Detección de odio en futbol en español

  • Montesinos-Cánovas, Esteban
  • García-Sánchez, Francisco
  • García-Díaz, José Antonio
  • Alcaraz Mármol, Gema
  • Valencia García, Rafael
In the last few years, Natural Language Processing (NLP) tools have been successfully applied to a number of different tasks, including author profiling, negation detection or hate speech detection, to name but a few. For the identification of hate speech from text, pre-trained language models can be leveraged to build high-performing classifiers using a transfer learning approach. In this work, we train and evaluate state-of-the-art pre-trained classifiers based on Transformers. The explored models are fine-tuned using a hate speech corpus in Spanish that has been compiled as part of this research. The corpus contains a total of 7,483 football-related tweets that have been manually annotated under four categories: aggressive, racist, misogynist, and safe. A multi-label approach is used, allowing the same tweet to be labeled with more than one class. The best results, with a macro F1-score of 88.713%, have been obtained by a combination of the models using Knowledge Integration.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Overview of FinancES 2023: Financial Targeted Sentiment Analysis in Spanish, Resumen de FinancES 2023: Análisis de Sentimiento Dirigido en Español sobre Finanzas

  • García-Díaz, José Antonio
  • Almela Sánchez-Lafuente, Ángela
  • García-Sánchez, Francisco
  • Alcaraz Mármol, Gema
  • Marín, María José
  • Valencia García, Rafael
This paper presents the FinancES 2023 shared task, organized in the IberLEF 2023 workshop, within the framework of the 39th International Conference of the Spanish Society for Natural Language Processing (SEPLN 2023). The aim of this task is to extend the challenge of sentiment analysis in Spanish to the financial domain, in order to extract the sentiment that a piece of financial information can have for several actors, including the main economic target (i.e., the specific company or asset where the economic fact applies), other companies (i.e., the entities producing the goods and services that others consume) and consumers (i.e., households/individuals). Specifically, two tasks are proposed and evaluated separately. One to identify the main target and to determine the sentiment polarity towards such target, and a second task to assess the sentiment towards both other companies and consumers. The ranking includes results for 10 different teams proposing novel approaches, mostly based on Transformers and generative language models.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Overview of PoliticEs 2022: Spanish Author Profiling for Political Ideology, Resumen de la tarea PoliticEs 2022: Perfilado del Autor Español por su Ideología Política

  • García-Díaz, José Antonio
  • Jiménez Zafra, Salud M.
  • Martín Valdivia, María Teresa
  • García-Sánchez, Francisco
  • Ureña López, Luis Alfonso
  • Valencia García, Rafael
This paper presents the PoliticEs 2022 shared task, organized at IberLEF 2022 workshop, within the framework of the 38th International Conference of the Spanish Society for Natural Language Processing. This task aims to extract the political ideology from a given user’s set of tweets. Specifically, it focused on the identification of the gender and the profession, as demographic traits, and the political ideology from a binary and multi-class perspective, as a psychographic trait. The PoliticEs task attracted 63 teams that registered through CodaLab. Finally, 20 submitted results and 14 presented working notes describing their systems. Most of the teams proposed transformer-based approaches, although some of them also used traditional machine learning algorithms or even a combination of both approaches.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Overview of PoliticES at IberLEF 2023: Political Ideology Detection in Spanish Texts, Resumen de la tarea PoliticES en IberLEF 2023: Detección de Ideología Política en Español

  • García-Díaz, José Antonio
  • Jiménez Zafra, Salud M.
  • Martín Valdivia, María Teresa
  • García-Sánchez, Francisco
  • Ureña López, Luis Alfonso
  • Valencia García, Rafael
This paper describes PoliticES 2023, a shared task organized within the workshop IberLEF 2023 in the framework of the 39th edition of the International Conference of the Spanish Society for Natural Language Processing. This second edition of the task shares the goal of the first edition of PoliticES, which is to extract political ideology and other psychographic and demographic characteristics of users in social networks. What is new this year is that the traits are extracted from text clusters of users who share the same traits, and that celebrities have been included as a type of profession. This edition attracted 43 teams, of which 11 submitted results and 8 sent papers describing their systems. Most of the participants proposed Transformers-based approaches, but others also used traditional machine learning algorithms.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Linguistic features integration for text classification tasks in Spanish, Integración de características lingüísticas para tareas de clasificación de texto en español

  • García-Díaz, José Antonio
This manuscript summarises the doctoral thesis of José Antonio García-Díaz at the University of Murcia, under the supervision of doctors Rafael Valencia-García and Pedro José Vivancos-Vicente. This doctoral thesis is published by compendium of publications under the industrial doctorate modality. The act of defence took place on Tuesday, July 5, 2022, before the court composed of doctors Salud María Jiménez-Zafra, from the University of Jaén; Miguel Ángel Rodríguez-García, from the Rey Juan Carlos University; and Mª del Pilar Salas-Zarate, from the National Technological Institute of Mexico. The qualification obtained was Outstanding Cum Laude unanimously. In addition, the mention of international doctorate was obtained.