Cargando...
Fecha
2025-06-22
Derechos de acceso
info:eu-repo/semantics/openAccess
Título de la revista
ISSN de la revista
Título del volumen
Editorial
Citas
Resumen
En un contexto de creciente mediatización del espacio político, los medios de comunicación actúan a la vez como seguidores del proceso parlamentario y, deliberadamente o no, como influencias activas sobre el desarrollo de los debates y prioridades políticas. Este trabajo aborda la cuestión de la correlación entre medios de comunicación y el proceso político, en el marco del paradigma del agenda setting, entendido como el proceso mediante el cual los medios contribuyen a definir qué temas adquieren relevancia pública y política en un determinado momento. Con frecuencia, este tipo de influencias se examinan mediante estudios cualitativos centrados en un caso de uso concreto, lo que exige un importante esfuerzo de análisis adaptado a cada situación. Frente a ello, se propone un enfoque metodológico que permita una medición sistemática y generalista de dicha influencia sobre temáticas extraídas directamente sobre el conjunto de canales de texto a estudiar. Tomando como objeto de estudio los debates parlamentarios y su posible correlación con los portales digitales de noticias, se propone un enfoque basado en el análisis de series temporales. La metodología propuesta parte de la clasificación de los textos en distintos niveles de representación, desde clasificaciones temáticas generales obtenidas mediante modelos de lenguaje pre-entrenados hasta extracciones más específicas de entidades y narrativas por modelos generativos; construyendo posteriormente las series temporales a partir de la aparición de estas etiquetas. Esto permite evaluar relaciones de influencia entre diferentes canales generadores de textos. Ya sea considerando como canal cada grupo parlamentario o portal de noticias; o considerando el discurso político y mediático al completo. El objetivo final es evaluar si, dada una temática, es posible observar patrones de influencia que resulten coherentes con las orientaciones ideológicas esperadas de los distintos canales. Partiendo de la hipótesis de que la influencia será más nítida cuanto más específica sea la temática o entidad analizada, lo que permitiría obtener una caracterización más fina de las dinámicas de agenda entre medios de comunicación y agrupaciones parlamentarias. La alta periodicidad del calendario legislativo y el contexto institucional del Congreso, suponen una limitación al compararlo con medios digitales de actualización constante. Aun así, se ha identificado una notable estabilidad en los temas tratados por el Congreso, salvo en los periodos vacacionales, y una diferencia clara en cómo prensa y política abordan temporalmente una misma temática. No se ha identificado una influencia generalizada de un portal digital sobre un partido concreto, aunque sí aparecen casos puntuales, de impacto limitado y sin persistencia clara entre temáticas.
In a context of increasing mediatization of the political space, the media act both as followers of the parliamentary process and, whether deliberately or not, as active influences on the development of political debates and priorities. This study addresses the correlation between media and the political process within the framework of agenda setting, defined as the process by which the media help to settle which topics gain public and political relevance at a given time. These types of influences are often examined through qualitative studies focused on one specific case study, which require a significant amount of analysis tailored to each situation. This work proposes a systematic and general methodological approach that allows for measuring influence across sets of topics extracted directly from the text channels being analyzed. Focusing on parliamentary debates and their possible correlation with digital news portals, this work applies a time series analysis approach. The proposed methodology begins by classifying the texts at different levels of representation, ranging from general topic classifications to more specific extractions of entities and narratives obtained through pre-trained language models, and then builds time series based on the occurrence of these labels. This allows for the evaluation of influence relationships between different text channels. Whether considering each parliamentary group and news portal as a channel, or considering the broad political and media discourse as a whole. The objetive is to evaluate whether, for a given topic, it is possible to observe patterns of influence that are consistent with the expected ideological orientations of the different channels. The hypothesis is that influence will become clearer as the topic or entity under analysis becomes more specific, allowing for a more detailed characterization of agenda dynamics between media outlets and parliamentary groups. The periodicity of the legislative calendar and the institutional context of Parliament pose a limitation when compared to the constantly updated nature of digital media. However, a notable stability has been identified in the topics addressed by Parliament, except during recess periods, and a clear difference in how the press and political parties approach the temporal evolution of the same themes. No generalized influence of a specific digital portal over political parties has been detected, although some isolated cases appear, with limited impact and no consistent pattern across topics.
In a context of increasing mediatization of the political space, the media act both as followers of the parliamentary process and, whether deliberately or not, as active influences on the development of political debates and priorities. This study addresses the correlation between media and the political process within the framework of agenda setting, defined as the process by which the media help to settle which topics gain public and political relevance at a given time. These types of influences are often examined through qualitative studies focused on one specific case study, which require a significant amount of analysis tailored to each situation. This work proposes a systematic and general methodological approach that allows for measuring influence across sets of topics extracted directly from the text channels being analyzed. Focusing on parliamentary debates and their possible correlation with digital news portals, this work applies a time series analysis approach. The proposed methodology begins by classifying the texts at different levels of representation, ranging from general topic classifications to more specific extractions of entities and narratives obtained through pre-trained language models, and then builds time series based on the occurrence of these labels. This allows for the evaluation of influence relationships between different text channels. Whether considering each parliamentary group and news portal as a channel, or considering the broad political and media discourse as a whole. The objetive is to evaluate whether, for a given topic, it is possible to observe patterns of influence that are consistent with the expected ideological orientations of the different channels. The hypothesis is that influence will become clearer as the topic or entity under analysis becomes more specific, allowing for a more detailed characterization of agenda dynamics between media outlets and parliamentary groups. The periodicity of the legislative calendar and the institutional context of Parliament pose a limitation when compared to the constantly updated nature of digital media. However, a notable stability has been identified in the topics addressed by Parliament, except during recess periods, and a clear difference in how the press and political parties approach the temporal evolution of the same themes. No generalized influence of a specific digital portal over political parties has been detected, although some isolated cases appear, with limited impact and no consistent pattern across topics.
Descripción
Categorías UNESCO
Palabras clave
Agenda setting, Series temporales, Modelos VAR, Medios digitales, Discurso parlamentario, Clasificación temática, Agenda setting, Time series, VAR models, Digital media, Parliamentary discourse, Topic classification
Citación
Pérez Miguel, Luis. Trabajo fin de Master: "¿Quién marca la agenda? Estudio de la influencia mediática en el discurso parlamentario mediante series temporales y clasificación textual multinivel" Universidad Nacional de Educación a Distancia (UNED), 2025
Centro
E.T.S. de Ingeniería Informática
Departamento
Lenguajes y Sistemas Informáticos