Procesamiento del Lenguaje Natural
http://journal.sepln.org/sepln/ojs/ojs/index.php/pln
<div class="homeText"><p>El objetivo principal de la revista es el de ofrecer a los investigadores en Procesamiento del Lenguaje Natural (PLN) una oportunidad para presentar nuevos trabajos, comunicar resultados, discutir problemas y obstáculos encontrados durante su trayectoria investigadora.</p><p>Por otro lado, permitir intercambiar opiniones sobre directrices futuras de investigación básica y aplicación prevista por los expertos y contrastarlas con las necesidades reales del mercado. Reflexionar y debatir en profundidad sobre temas concretos de máxima actualidad tales como la extracción de información, la recuperación de información o la evaluación de sistemas de procesamiento del lenguaje natural.</p><p>La Revista tiene una periodicidad semestral, publicándose dos números al año (marzo y septiembre) que recogen los últimos avances en PLN.</p><p>La Revista cuenta con el sello de calidad de la Fundación Española para Ciencia y Tecnología (FECyT), el cual la certifica como revista de excelencia, y por lo tanto, incluida en el Repositorio de Revistas Científicas españolas (RECyT, <span>Repositorio Español de Ciencia y Tecnología) <a href="http://recyt.fecyt.es/index.php/PLN">http://recyt.fecyt.es/index.php/PLN</a></span></p><p><span>La Revista de Procesamiento de Lenguaje Natural también ha recibido e<span>l sello de calidad (ISO9001) que la acredita como excelente durante un periodo de tres años (14 de marzo de 2012 al 14 de marzo de 2015).</span></span></p><p>Procesamiento del Lenguaje Natural (edición impresa). ISSN: 1135-5948.</p><p>Procesamiento del Lenguaje Natural (edición electrónica). ISSN: 1989-7553.</p></div>Sociedad Española para el Procesamiento del Lenguaje Naturales-ESProcesamiento del Lenguaje Natural1135-5948Overview of Rest-Mex at IberLEF 2023: Research on Sentiment Analysis Task for Mexican Tourist Texts
http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/6572
This paper presents the framework and results of the Rest-Mex task at IberLEF 2023, focusing on sentiment analysis and text clustering of tourist texts. The study primarily focuses on texts related to tourist destinations in Mexico, although this edition included data from Cuba and Colombia for the first time. The sentiment analysis task aims to predict the polarity of opinions expressed by tourists, classifying the type of place visited, whether it's a tourist attraction, hotel, or restaurant, as well as the country it is located in. On the other hand, the text clustering task aims to classify news articles related to tourism in Mexico. For both tasks, corpora were built using Spanish opinions extracted from TripAdvisor and news articles from Mexican media. This article compares and discusses the results obtained by the participants in both sub-tasks. Additionally, a method is proposed to measure the easiness of a multi-class text classification corpus, along with an approach for system selection in a possible late fusion scheme.Miguel Ángel Álvarez-CarmonaÁngel Díaz-PachecoRamón ArandaAnsel Yoan Rodríguez-GonzálezVictor Muñiz-SánchezAdrián Pastor López-MonroyFernando Sánchez-VegaLázaro Bustio-Martínez
Copyright (c) 2023 Procesamiento del Lenguaje Natural
2023-09-132023-09-1371425436Overview of FinancES 2023: Financial Targeted Sentiment Analysis in Spanish
http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/6571
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.José Antonio Garcia-DíazÁngela AlmelaFrancisco García-SánchezGema Alcaraz-MármolMaría José MarínRafael Valencia-García
Copyright (c) 2023 Procesamiento del Lenguaje Natural
2023-09-132023-09-1371417423Overview of PoliticES at IberLEF 2023: Political Ideology Detection in Spanish Texts
http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/6570
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.José Antonio Garcia-DíazSalud María Jiménez-ZafraMaría-Teresa Martín-ValdiviaFrancisco García-SánchezLuis Alfonso Ureña-LópezRafael Valencia-García
Copyright (c) 2023 Procesamiento del Lenguaje Natural
2023-09-132023-09-1371409416Overview of DIPROMATS 2023: automatic detection and characterization of propaganda techniques in messages from diplomats and authorities of world powers
http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/6569
This paper presents the results of the DIPROMATS 2023 challenge, a shared task included at the Iberian Languages Evaluation Forum (IberLEF). DIPROMATS 2023 provides a dataset with 12012 annotated tweets in English and 9501 tweets in Spanish, posted by authorities of China, Russia, United States and the European Union. Three tasks are proposed for each language. The first one aims to distinguish if a tweet has propaganda techniques or not. The second task seeks to classify the tweet into four clusters of propaganda techniques, whereas the third one offers a fine-grained categorization of 15 techniques. For the three tasks we have received a total of 34 runs from 9 different teamsPablo MoralGuillermo MarcoJulio GonzaloJorge Carrillo-de-AlbornozIván Gonzalo-Verdugo
Copyright (c) 2023 Procesamiento del Lenguaje Natural
2023-09-132023-09-1371397407Everybody Hurts, Sometimes Overview of HUrtful HUmour at IberLEF 2023: Detection of Humour Spreading Prejudice in Twitter
http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/6568
Humour is an efficient strategy to spread prejudice because, most of the time, it evades moral judgement. However, it perpetuates stereotypes and doing so justifies discriminatory acts. At HUHU we propose a frame to study how humour is used to discriminate minorities and to analyse their interplay with the degree of prejudice expressed against specific groups. To this end, we provide a corpus of prejudiced tweets in Spanish annotated with the presence of humour, its prejudice degree and the targeted groups: women and feminists, the LGBTI+ community, immigrants and racially discriminated people, and over-weighted people. This paper analyses the results achieved by the 46 teams that participated in HUHU.Roberto Labadie TamayoBerta ChulviPaolo Rosso
Copyright (c) 2023 Procesamiento del Lenguaje Natural
2023-09-132023-09-1371383395Overview of HOPE at IberLEF 2023: Multilingual Hope Speech Detection
http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/6567
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.Salud María Jiménez-ZafraMiguel Ángel Garcia-CumbrerasDaniel García-BaenaJosé Antonio Garcia-DíazBharathi Raja ChakravarthiRafael Valencia-GarcíaLuis Alfonso Ureña-López
Copyright (c) 2023 Procesamiento del Lenguaje Natural
2023-09-132023-09-1371371381Overview of HOMO-MEX at Iberlef 2023: Hate speech detection in Online Messages directed Towards the MEXican Spanish speaking LGBTQ+ population
http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/6566
The detection of hate speech and stereotypes in online platforms has gained significant attention in the field of Natural Language Processing (NLP). Among various forms of discrimination, LGBTQ+ phobia is prevalent on social media, particularly on platforms like Twitter. The objective of the HOMO-MEX task is to encourage the development of NLP systems that can detect and classify LGBTQ+ phobic content in Spanish tweets, regardless of whether it is expressed aggressively or subtly. The task aims to address the lack of dedicated resources for LGBTQ+ phobia detection in Spanish Twitter and encourages participants to employ multi-label classification approaches.Gemma Bel-EnguixHelena Gómez-AdornoGerardo SierraJuan VásquezScott Thomas AndersenSergio Ojeda-Trueba
Copyright (c) 2023 Procesamiento del Lenguaje Natural
2023-09-132023-09-1371361370Overview of DA-VINCIS at IberLEF 2023: Detection of Aggressive and Violent Incidents from Social Media in Spanish
http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/6565
In this paper, we present the overview of the DA-VINCIS 2023 shared task which was organized at IberLEF 2023 and co-located in the framework of the 39th International Conference of the Spanish Society for Natural Language Processing (SEPLN 2023). The main aim of this task is to promote the research on developing automatic solutions for detecting violent events in social networks. Two subtasks were considered: (i) A binary classification task aimed to determine whether or not a tweet is about a violent incident; and (ii) A multi-label multi-class classification task in which the category(ies) of a violent incident must be identified. A multimodal manual annotated corpus comprising both tweets and images associated to them was provided to the participants. A total of 15 systems were submitted for the final evaluation phase. Competitive results were obtained for both subtasks, the higher ones were in the binary classification task. Corpora and results are available at the shared task website at https://codalab.lisn.upsaclay.fr/competitions/11312.Horacio Jarquín-VásquezDelia Irazú Hernández-FaríasLuis Joaquín ArellanoHugo Jair EscalanteLuis Villaseñor-PinedaManuel Montes-y-GómezFernando Sanchez-Vega
Copyright (c) 2023 Procesamiento del Lenguaje Natural
2023-09-132023-09-1371351360Overview of MentalRiskES at IberLEF 2023: Early Detection of Mental Disorders Risk in Spanish
http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/6564
This paper presents the MentalRiskEs shared task organized at IberLEF 2023, as part of the 39th International Conference of the Spanish Society for Natural Language Processing (SEPLN 2023). The aim of this task is to promote the early detection of mental risk disorders in Spanish. We outline three detection tasks: Task 1 on eating disorders, Task 2 on depression, and Task 3 on an undisclosed disorder during the competition (anxiety) to observe the transfer of knowledge among the different disorders proposed. Furthermore, we asked participants to submit measurements of carbon emissions for their systems, emphasizing the need for sustainable NLP practices. In this first edition, 37 teams registered, 18 submitted results, and 16 presented papers. Most teams experimented with Transformers, including features, data augmentation, and preprocessing techniques.Alba María Mármol-RomeroAdrián Moreno-MuñozFlor Miriam Plaza-del-ArcoMaría Dolores Molina-GonzálezMaria Teresa Martín-ValdiviaLuis Alfonso Ureña-LópezArturo Montejo-Raéz
Copyright (c) 2023 Procesamiento del Lenguaje Natural
2023-09-132023-09-1371329350Overview of GUA-SPA at IberLEF 2023: Guarani-Spanish Code Switching Analysis
http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/6563
We present the first shared task for detecting and analyzing codeswitching in Guarani and Spanish, GUA-SPA at IberLEF 2023. The challenge consisted of three tasks: identifying the language of a token, NER, and a novel task of classifying the way a Spanish span is used in the code-switched context. We annotated a corpus of 1500 texts extracted from news articles and tweets, around 25 thousand tokens, with the information for the tasks. Three teams took part in the evaluation phase, obtaining in general good results for Task 1, and more mixed results for Tasks 2 and 3.Luis ChiruzzoMarvin Agüero-ToralesGustavo Giménez-LugoAldo AlvarezYliana RodríguezSantiago GóngoraThamar Solorio
Copyright (c) 2023 Procesamiento del Lenguaje Natural
2023-09-132023-09-1371321328