Análisis de los errores lingüísticos generados por la IA en el doblaje automático de expresiones idiomáticas deportivas en programas de televisión en vivo.
Loading...
Date
2025
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad de Concepción
Abstract
En el presente estudio se analizaron los errores producidos por la inteligencia artificial (IA) al doblar de manera automática expresiones idiomáticas en una transmisión de un partido de fútbol en vivo, mediante un enfoque descriptivo-analítico. Para ello, se recopilaron y estudiaron expresiones extraídas del partido MLS NEXT Pro: Orlando City B vs. Inter Miami II y, posteriormente, se clasificaron según el tipo de error cometido por la IA en la traducción y tipología idiomática con su respectivo grado de equivalencia. Los resultados revelaron que la IA presenta dificultades recurrentes en la elección léxica y la conservación del sentido figurado, lo que genera traducciones con carencia de naturalidad y fluidez. Asimismo, se observó que las expresiones idiomáticas semi-puras y puras presentan problemas al ser traducidas automáticamente y que la mayoría de las propuestas solo ofrecen una equivalencia parcial. Estos hallazgos evidencian que, si bien la IA es una herramienta de gran utilidad que traduce correctamente las expresiones con menor carga idiomática, aún no logra reproducir la naturalidad y expresividad propias del relato futbolístico, lo que reafirma la relevancia del traductor humano en el proceso del doblaje deportivo.
This study analyzed the errors produced by artificial intelligence (AI) when automatically translating idiomatic expressions in a live soccer match broadcast, using a descriptive-analytical approach. Expressions were collected from the MLS NEXT Pro: Orlando City B vs. Inter Miami II match and subsequently classified according to the type of error made by AI in translation, as well as their idiomatic typology and corresponding degree of equivalence. The results revealed that AI has recurring difficulties in lexical choice and the preservation of figurative meaning, which generates translations that lack naturalness and fluency. Likewise, it was observed that semi-pure and pure idiomatic expressions present problems when translated automatically and that most of the proposals only offer partial equivalence. These findings show that, although AI can be a useful tool for correctly translating expressions with a lower idiomatic load, it still fails to reproduce the naturalness and expressiveness characteristic of football commentary, thereby reaffirming the importance of the human translator in the process of sports dubbing.
This study analyzed the errors produced by artificial intelligence (AI) when automatically translating idiomatic expressions in a live soccer match broadcast, using a descriptive-analytical approach. Expressions were collected from the MLS NEXT Pro: Orlando City B vs. Inter Miami II match and subsequently classified according to the type of error made by AI in translation, as well as their idiomatic typology and corresponding degree of equivalence. The results revealed that AI has recurring difficulties in lexical choice and the preservation of figurative meaning, which generates translations that lack naturalness and fluency. Likewise, it was observed that semi-pure and pure idiomatic expressions present problems when translated automatically and that most of the proposals only offer partial equivalence. These findings show that, although AI can be a useful tool for correctly translating expressions with a lower idiomatic load, it still fails to reproduce the naturalness and expressiveness characteristic of football commentary, thereby reaffirming the importance of the human translator in the process of sports dubbing.
Description
Tesis presentada para optar al grado de Licenciado en Traductología.
Keywords
Traducción e interpretación, Inteligencia artificial, Doblaje de programas de televisión