The integration of Large Language Models (LLMs) in education requires educators to develop advanced digital competencies that encompass technical, pedagogical, and ethical dimensions. This study presents an analysis of a comprehensive training initiative involving the design and implementation of a specialized course on LLM personalization for academic instructors, vocational trainers, corporate learning and development managers, as well as designers of educational and training materials. This experience includes the development of a software tool, the ProyectoHeliox LLMeducakit, an open-source platform based on OpenWebUI and tailored for educators to ease the LLM personalization process and management. We propose a systematic analysis of student projects of this course to assess advanced digital competences development. Results reveal six principal categories of advanced LLM personalization options in educational contexts, each fostering skills such as computational thinking, ethical AI use, data management, and pedagogical innovation. The features of LLMeducakit, including accessible, privacy-preserving, and customizable AI workflows, were instrumental in supporting these advancements. The findings demonstrate that a combination of targeted professional development and dedicated technological tools enables trainers to move from passive AI users to active co-designers of AI-driven educational solutions, aligning with international frameworks and highlighting the importance of ethical, inclusive, and reflective AI integration in education.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Empowering Educators as AI Co-designers: Developing Advanced Digital Competencies with a Customized LLMeducakit for Large Language Model Personalization

  • Roberto Feltrero,
  • Luis-F. López-López

摘要

The integration of Large Language Models (LLMs) in education requires educators to develop advanced digital competencies that encompass technical, pedagogical, and ethical dimensions. This study presents an analysis of a comprehensive training initiative involving the design and implementation of a specialized course on LLM personalization for academic instructors, vocational trainers, corporate learning and development managers, as well as designers of educational and training materials. This experience includes the development of a software tool, the ProyectoHeliox LLMeducakit, an open-source platform based on OpenWebUI and tailored for educators to ease the LLM personalization process and management. We propose a systematic analysis of student projects of this course to assess advanced digital competences development. Results reveal six principal categories of advanced LLM personalization options in educational contexts, each fostering skills such as computational thinking, ethical AI use, data management, and pedagogical innovation. The features of LLMeducakit, including accessible, privacy-preserving, and customizable AI workflows, were instrumental in supporting these advancements. The findings demonstrate that a combination of targeted professional development and dedicated technological tools enables trainers to move from passive AI users to active co-designers of AI-driven educational solutions, aligning with international frameworks and highlighting the importance of ethical, inclusive, and reflective AI integration in education.