Abstract:

Education is going through a progressive transformation with the rise of artificial intelligence. A wide range of Generative Artificial Intelligence (GenAI) tools have recently emerged that are used to perform diverse activities in the educational context. In this field, Large Language Models (LLMs) are mainly focusing on three lines of academic research: Study Assisting, Teach Assisting, and Adaptive Learning. In the programming education field, these models have undergone a remarkable transformation in the creation of material, through generating code fragments in a short period of time. However, given the tendency of LLMs to generate inconsistent results, their effective implementation requires expert human supervision to ensure the technical accuracy required for learning.

A fundamental approach to enhance learning is through personalization. Nowadays, technological advances have enabled teachers to identify students' individual features, such as their personality, learning style, or emotional state, therefore allowing the adaptation of study materials to their particular needs. This focus does not only support the teacher in their educational work but also promotes an environment in which each student's individual needs can effectively and promptly be addressed.

The main purpose of this article is to describe the pedagogical integration of GenAI in adaptive learning environments, using the COVIA environment as a reference. COVIA is an environment that uses video-based learning and a computational notebook to aid students in the development of algorithmic logic and program coding using the Java programming language as a basis. The integration of GenAI also aims to reduce the teacher´s workload associated with modifying and generating educational resources. To address this issue, two modules that use GenAI as the foundation for their operation were implemented. These modules include the personalization of exercises and a chatbot integration to assist students in exercises resolution according to their learning style and other factors.

Keywords:

Generative Artificial Intelligence, Large Language Models, Personalized Education, Programming, Chatbot, Educational Resources.