The Effective Use of Generative AI for Personalized Learning
Chapter
Published version
Permanent lenke
https://hdl.handle.net/11250/3173160Utgivelsesdato
2024Metadata
Vis full innførselSamlinger
- Institutt for design [1215]
- Publikasjoner fra CRIStin - NTNU [39811]
Sammendrag
This study explored the possibilities of using generative artificial intelligence (GAI) for text adaptation based on a created persona, i.e., a 12-year-old student with reading comprehension challenges caused by ADHD. The process entailed comparing three outputs generated in three different ways: “out-of-box” use of GAI, GAI-teacher collaboration, and teachers only. Each generated adapted text was evaluated by another group of teachers to understand their suitability for the persona. The results showed that the teacher-adapted text received the highest score, followed by the “out-of-the-box” adapted text and the GAI-teacher collaborative text. The resulting texts utilized different adaptation methods to fit the content towards the persona, engaging the reader, improving the structure, and adjusting the difficulty of the text. However, GAI could be a valid option for personalized learning through text adaptation, considering the close scores between the adapted texts. With the effective use of prompt design, we believe it could assist teachers in adapting content and save them valuable time.