Persona: Benito Santos, Alejandro
Cargando...
Dirección de correo electrónico
al.benito@lsi.uned.es
ORCID
0000-0001-5317-6390
Fecha de nacimiento
Proyectos de investigación
Unidades organizativas
Puesto de trabajo
Apellidos
Benito Santos
Nombre de pila
Alejandro
Nombre
5 resultados
Resultados de la búsqueda
Mostrando 1 - 5 de 5
Publicación Test-driving information theory-based compositional distributional semantics: A case study on Spanish song lyrics(ELSEVIER, 2025-06-15) Ghajari Espinosa, Adrián; Benito Santos, Alejandro; Ros Muñoz, Salvador; Fresno Fernández, Víctor Diego; González Blanco, ElenaSong lyrics pose unique challenges for semantic similarity assessment due to their metaphorical language, structural patterns, and cultural nuances - characteristics that often challenge standard natural language processing (NLP) approaches. These challenges stem from a tension between compositional and distributional semantics: while lyrics follow compositional structures, their meaning depends heavily on context and interpretation. The Information Theory-based Compositional Distributional Semantics framework offers a principled approach by integrating information theory with compositional rules and distributional representations. We evaluate eight embedding models on Spanish song lyrics, including multilingual, monolingual contextual, and static embeddings. Results show that multilingual models consistently outperform monolingual alternatives, with the domain-adapted ALBERTI achieving the highest F1 macro scores (78.92 ± 10.86). Our analysis reveals that monolingual models generate highly anisotropic embedding spaces, significantly impacting performance with traditional metrics. The Information Contrast Model metric proves particularly effective, providing improvements up to 18.04 percentage points over cosine similarity. Additionally, composition functions maintaining longer accumulated vector norms consistently outperform standard averaging approaches. Our findings have important implications for NLP applications and challenge standard practices in similarity calculation, showing that effectiveness varies with both task nature and model characteristics.Publicación Characterizing the visualization design space of distant and close reading of poetic rhythm(Frontiers, 2023-06-06) Benito Santos, Alejandro; Ros Muñoz, Salvador; Therón Sánchez, Roberto; García Peñalvo, Francisco J.; Agencia Estatal de Investigación (España)Metrical and rhythmical poetry analysis is founded on the systematic statistical analysis and comparison of sonic devices (e.g., rhythmic patterns) that emerge from a combination of pre-established aesthetic and structural rules and the poet's abilities and creative genius to convey a given message adhering to the said constraints. These rhythmical patterns, which have been traditionally obtained by means of a careful close reading of the poems, in a process known as “scansion,” can now be obtained and made visible by automatic means. However, the visualization literature is still scarce on approaches that allow an insightful close and distant reading of the rhythmical patterns in a poetry corpus. In this work, we report our initial efforts in characterizing of the visualization design space of distant and close reading of poetic rhythm. By employing a digital version of a corpus of 11,268 verses originally written by the Spanish poet and playwright Federico García-Lorca (1898–1936), we could craft several prototypical visualizations representative of the inherent complexity of the problem which we expect to employ in future user studies and that we share here with the rest of the community to foster further discussion around this interesting topic.Publicación Robust Estimation of Population-Level Effects in Repeated-Measures NLP Experimental Designs(Association for Computational Linguistics, 2025-01-01) Benito Santos, Alejandro; Ghajari Espinosa, Adrián; Fresno Fernández, Víctor DiegoNLP research frequently grapples with multiple sources of variability—spanning runs, datasets, annotators, and more—yet conventional analysis methods often neglect these hierarchical structures, threatening the reproducibility of findings. To address this gap, we contribute a case study illustrating how linear mixed-effects models (LMMs) can rigorously capture systematic language-dependent differences (i.e., population-level effects) in a population of monolingual and multilingual language models. In the context of a bilingual hate speech detection task, we demonstrate that LMMs can uncover significant population-level effects—even under low-resource (small-N) experimental designs—while mitigating confounds and random noise. By setting out a transparent blueprint for repeated-measures experimentation, we encourage the NLP community to embrace variability as a feature, rather than a nuisance, in order to advance more robust, reproducible, and ultimately trustworthy results.Publicación BKViz: A Basketball Visual Analysis Tool(Institute of Electrical and Electronics Engineers (IEEE), 2016-11-21) Losada Gómez, Antonio G.; Therón Sánchez, Roberto; Benito Santos, AlejandroThe amount of data available nowadays in the sports eld is hard to comprehend using classic analytic methods. This calls for the development of systems such as the prototype discussed here, which makes it possible to manipulate chunks of data to then portray them in visual ways, easing their understanding. Based on basketball, this tool helps users in reaching conclusions regarding performances during individual matches. This enables them to gather knowledge about play sequences, the events occurring at di erent moments, and the style of play teams employ based on player chemistry. Using multiple visualizations in an integrated system, data can be manipulated to vary how they are shown and change their analytic power as needed. With interaction and simple comprehension based on visualization being the two cornerstones, the challenge was to provide a tool able to showcase patterns and study actions in an innovative and combined way not yet exploited.Publicación Playing Design: A Case Study on Applying Gamification to Construct a Serious Game with Youngsters at Social Risk(Association for Computing Machinery (ACM), 2021-04-23) Benito Santos, Alejandro; Losada Gómez, Antonio G.; Palfinger, Thomas; Therón Sánchez, Roberto; Wandl-Vogt, EvelineThis article reports on the experience of co-designing an educational video game aimed at promoting good dietary habits in youngsters and fostering Sustainable Development Goals (SDGs), such as SDG 3 (Good Health and Well-Being), SDG 10 (Reduced Inequalities), and SDG 17 (Partnerships for the Goals). To ensure the quality of the results, we developed a methodology under a social innovation paradigm that enabled the co-creation of the game. The methodology was driven by a series of three workshops, during which we adopted several different gamification strategies to support a Participatory Design (PD) process with the stakeholders, a group of local pre-teen and teen girls at social risk (N = 22). Captured requirements materialized into intermediate prototype evaluations that motivated a progressive refinement of the game.