Pasar al contenido principal

BEWORD-UPV: DESCUBRIENDO EL SIGNIFICADO Y LA INTENCION MAS ALLA DE LA PALABRA HABLADA: HACIA UN ENTORNO INTELIGENTE PARA ABORDAR LOS DOCUMENTOS MULTIMEDIA

PID2021-126061OB-C41

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal para Impulsar la Investigación Científico-Técnica y su Transferencia
Subprograma Subprograma Estatal de Generación de Conocimiento
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2021
Plan Estatal 2021-2023 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSITAT POLITÈCNICA DE VALÈNCIA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 5
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Attentional Extractive Summarization

  • González-Barba, José Ángel
  • Segarra Soriano, Encarnación
  • García-Granada, Fernando|||0000-0003-2213-4213
  • Sanchís Arnal, Emilio|||0000-0002-6737-4723
  • Hurtado Oliver, Lluis Felip|||0000-0002-1877-0455
[EN] In this work, a general theoretical framework for extractive summarization is proposed¿the Attentional Extractive Summarization framework. Although abstractive approaches are generally used in text summarization today, extractive methods can be especially suitable for some applications, and they can help with other tasks such as Text Classification, Question Answering, and Information Extraction. The proposed approach is based on the interpretation of the attention mechanisms of hierarchical neural networks, which compute document-level representations of documents and summaries from sentence-level representations, which, in turn, are computed from word-level representations. The models proposed under this framework are able to automatically learn relationships among document and summary sentences, without requiring Oracle systems to compute the reference labels for each sentence before the training phase. These relationships are obtained as a result of a binary classification process, the goal of which is to distinguish correct summaries for documents. Two different systems, formalized under the proposed framework, were evaluated on the CNN/DailyMail and the NewsRoom corpora, which are some of the reference corpora in the most relevant works on text summarization. The results obtained during the evaluation support the adequacy of our proposal and suggest that there is still room for the improvement of our attentional framework., This work is partially supported by MCIN/AEI/10.13039/501100011033 and by the European Union "NextGenerationEU/PRTR" under grants PDC2021-120846-C44 (AMIC-PoC-UPV) and PID2021-126061OB-C41 (BEWORD-UPV).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Abstractive Summarizers Become Emotional on News Summarization

  • Ahuir-Esteve, Vicent|||0000-0001-5636-651X
  • Hurtado Oliver, Lluis Felip|||0000-0002-1877-0455
  • González, José-Ángel
  • Segarra Soriano, Encarnación
[EN] Emotions are central to understanding contemporary journalism; however, they are overlooked in automatic news summarization. Actually, summaries are an entry point to the source article that could favor some emotions to captivate the reader. Nevertheless, the emotional content of summarization corpora and the emotional behavior of summarization models are still unexplored. In this work, we explore the usage of established methodologies to study the emotional content of summarization corpora and the emotional behavior of summarization models. Using these methodologies, we study the emotional content of two widely used summarization corpora: Cnn/Dailymail and Xsum, and the capabilities of three state-of-the-art transformer-based abstractive systems for eliciting emotions in the generated summaries: Bart, Pegasus, and T5. The main significant findings are as follows: (i) emotions are persistent in the two summarization corpora, (ii) summarizers approach moderately well the emotions of the reference summaries, and (iii) more than 75% of the emotions introduced by novel words in generated summaries are present in the reference ones. The combined use of these methodologies has allowed us to conduct a satisfactory study of the emotional content in news summarization., This work is partially supported by MCIN/AEI/10.13039/501100011033, by the European Union and NextGenerationEU/MRR , and by ERDF A way of making Europe under grants PDC2021-120846-C44 and PID2021-126061OB-C41. It is also partially supported by the Generalitat Valenciana under project CIPROM/2021/023, and by the Spanish Ministerio de Universidades under the grant FPU21/05288 for university teacher training.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Journalists&apos, Ethical Responsibility: Tackling Hate Speech Against Women Politicians in Social Media Through Natural Language Processing Techniques

  • Iranzo-Cabrera, M.
  • Simon-Astudillo, I.
  • Castro-Bleda, Maria Jose|||0000-0003-1001-8258
  • Hurtado Oliver, Lluis Felip|||0000-0002-1877-0455
[EN] Social media has led to a redefinition of the journalist's role. Specifically on Twitter, these professionals assume an influential position and their discourse is dominated by personal opinions. Taking into consideration that this platform has proven to be a breeding ground for polarization, digital harassment and hate speech, notably against women politicians, this research aims to analyze journalists' involvement in this complex scenario. The investigation aims to determine whether, immersed in online and gender defamation campaigns, journalists enhance the quality of public debate or, on the contrary, they reinforce the visibility of this hostile content. To this end, we examined a sample of 63,926 tweets published from 23 to 25 November 2022 related to a campaign of political violence against the Spanish Minister of Equality using Natural Language Processing tools and qualitative content analysis. Results show that during those three days, at least half of the tweets contained hate speech and improper language. In this climate of hostility, journalists participating in the debate not only have an ability to attract likes and retweets but also exhibit polarization and use hate speech. Each ideological position-for and against the Minister-is also reflected in their own uncivil strategies. Under the umbrella of free speech and regardless of argumentative discourses, those journalists who lean towards ideological progressivism tend to insult their opponents, and those on the political right use divisive constructions, stereotyping and irony as attack techniques., This work was supported by two Ministerio de Ciencia, Innovacion y Universidades (Spain) and ERDF A way of making Europe (PID2021-126061OB-C41) and Ministerio de Ciencia e Innovacion (PID2020-113574RB-I00).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Leveraging Transformers-based models and linked data for deep phenotyping in radiology

  • Hurtado Oliver, Lluis Felip|||0000-0002-1877-0455
  • MARCO RUIZ, LUIS|||0000-0001-6349-3162
  • Castro-Bleda, Maria Jose|||0000-0003-1001-8258
  • DE LA IGLESIA VAYÁ, MARIA DE LOS DESAMPARADOS
  • Segarra Soriano, Encarnación
  • Bustos-Moreno, Aurelia
  • Vallalta-Rueda, Juan Francisco
[EN] Background and Objective: Despite significant investments in the normalization and the standardization of Electronic Health Records (EHRs), free text is still the rule rather than the exception in clinical notes. The use of free text has implications in data reuse methods used for supporting clinical research since the query mechanisms used in cohort definition and patient matching are mainly based on structured data and clinical terminologies. This study aims to develop a method for the secondary use of clinical text by: (a) using Natural Language Processing (NLP) for tagging clinical notes with biomedical terminology; and (b) designing an ontology that maps and classifies all the identified tags to various terminologies and allows for running phenotyping queries. Methods and Results: Transformers-based NLP Models, concretely pre-trained RoBERTa language models, were used to process radiology reports and annotate them identifying elements matching UMLS Concept Unique Identifiers (CUIs) definitions. CUIs were mapped into several biomedical ontologies useful for phenotyping (e.g., SNOMED-CT, HPO, ICD-10, FMA, LOINC, and ICPC2, among others) and represented as a lightweight ontology using OWL (Web Ontology Language) constructs. This process resulted in a Linked Knowledge Base (LKB), which allows running expressive queries to retrieve reports that comply with specific criteria using automatic reasoning. Conclusion: Although phenotyping tools mostly rely on relational databases, the combination of NLP and Linked Data technologies allows us to build scalable knowledge bases using standard ontologies from the Web of data. Our approach enables us to execute a pipeline which input is free text and automatically maps identified entities to a LKB that allows answering phenotyping queries. In this work, we have only used Spanish radiology reports, although it is extensible to other languages for which suitable corpora are available. This is particularly valuable in regional and national systems dealing with large research databases from different registries and cohorts and plays an essential role in the scalability of large data reuse infrastructures that require indexing and governing distributed data sources., This work has been carried out during phase I of the Big Data Personalized Medicine project, file 18/CPP/1. The Big Data Personal-ized Medicine project has been co-financed by the European Regional Development Fund (FEDER) in 85% for the (sic)3,833,774 budgeted for the Canary Islands Health Service, and 50% for the (sic)2,000,000 budgeted for the Generalitat Valenciana, through a grant granted by the Ministry of Science and Innovation of (sic)4,258,707.90. It is also funded by the Ministerio de Ciencia e Innovacion, Spain and European Union under project BEWORD PID2021-126061OB-C41, and by the Generalitat Valenciana, Spain under projects CIPROM/2021/023 and PROMETEO/2020/024.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Digitization and recognition of Jacquard cards for textile design preservation

  • Marc Rodas Lorente
  • Robes Martin, Elena
  • España Boquera, Salvador|||0000-0002-0335-9997
  • Castro-Bleda, Maria Jose|||0000-0003-1001-8258
[EN] This paper introduces an advanced, structured approach to preserving textile design, specifically targeting the intricate silk motifs woven using Jacquard machines in the textile industry. Traditional Jacquard machines utilize perforated cards, where specific hole patterns guide needle movements to create complex designs. Due to deterioration over time, these cards need replacement to maintain design integrity, making digitization essential for preserving cultural heritage. To address this need, we developed a multi-step, computer vision-based method that extracts design information from images of these Jacquard cards. The process involves 1) segmenting each card from images, 2) identifying and locating hole patterns, 3) refining hole detection through interpolation and statistical validation, and 4) post-processing for error detection and correction. This results in highly accurate, vectorized digital files, which can be replicated using laser technology to produce new cards. Additionally, an algorithm generates corresponding fabric patterns directly from the digital designs. This systematic approach not only facilitates the preservation and replication of historic silk motifs but also supports ongoing innovation in textile design., We are grateful for the financial support from by ValgrAI - Valencian Graduate School and Research Network of Artificial Intelligence and the Generalitat Valenciana, the Spanish Ministerio de Ciencia e Innovación and European Union under project BEWORD PID2021-126061OB-C41, and by the Generalitat Valenciana under project PROMETEO/2020/024. Funding for open access charge: CRUE-Universitat Politècnica de València.