Sistemas recomendadores para el desarrollo de objetos de aprendizaje para educación en ingenierías: una revisión sistemática

This research presents a systematic review with the objective of identifying recommender systems designed to facilitate the creation of learning objects within higher education, with a particular focus on engineering education. The investigation delves into various facets including the recommended granularity level, filtering strategies, the utilization of artificial intelligence techniques, evaluation methodologies applied, pedagogical guidelines governing recommendation generation, and the integration of competency-based training considerations. To accomplish this, we extracted 409 initial references published between 2000 and 2023 from indexed journals. After a preliminary screening based on abstracts, 8 promising references were selected. After closer examination, only 3 of them proved relevant in addressing our research inquiries. Furthermore, an extended search on Google Scholar given as result 29 supplementary works that complemented the primary studies selected. Our analysis, guided by the research questions, revealed a dearth of recommender systems for learning object development. Despite the application of diverse artificial intelligence techniques, none of the identified systems recommend design at the medium level of granularity or take into account the competency-based approach. Furthermore, these systems are not helpful for educators without experience in instructional design. This finding reveals the necessity for further research that we commit to publish in RED once we finish it. Este trabajo presenta una revisión sistemática para determinar si existen sistemas recomendadores que den soporte al diseño de objetos de aprendizaje en educación superior, fundamentalmente para enseñanza de ingenierías. Se pretende conocer la granularidad recomendada, las estrategias de filtrado y técnicas de inteligencia artificial utilizadas, los métodos de evaluación aplicados, y la consideración del enfoque de competencias y lineamientos didáctico-pedagógicos para generar recomendaciones. Para ello, se recabaron 409 referencias iniciales publicadas entre 2000 y 2023 en revistas indexadas, de las cuales se preseleccionaron 8 según sus resúmenes. Luego de una lectura crítica quedaron 3, más 1 trabajo que se tenía previamente, para responder las preguntas de investigación. Una búsqueda adicional en Google Académico arrojó 29 trabajos complementarios a los estudios primarios seleccionados. Las preguntas de investigación guiaron la extracción y análisis de datos.