Using genre analysis to detect AIGenerated academic texts
Fecha
2024-07
Autores
Melliti, Mimoun
Título de la revista
ISSN de la revista
Título del volumen
Editor
Editorial Universidad Don Bosco
Resumen
This study investigates the distinguishing characteristics between human-written and AI-generated abstracts through genre analysis techniques. The research examined mini-memoir abstracts authored by MA2 students at Faculty of Arts and Humanities, University of Kairouan, Tunisia and compared them to AI-generated abstracts created specifically for this study using ChatGPT. The analysis focused on text function recurrence, specifically the frequency and quality of elements such as purpose statements, methodology, results, and contextualization. Findings revealed that human-written abstracts exhibit a more comprehensive and detailed presentation, emphasizing contextualization and thorough results, while AI-generated abstracts tend to prioritize clear and explicit purpose statements with less depth in results and contextual information. The study highlights the need for targeted teacher training and rigorous assessment criteria to uphold academic integrity and address the challenges posed by AI in scholarly writing.
Descripción
Palabras clave
Genre Analysis , AI-generated Texts , Academic Abstracts , Human-AI Comparison
Citación
Melliti, M. (2024). Using genre analysis to detect AI-Generated academic texts. Diá-Logos, 16(29), 09–27.