Please use this identifier to cite or link to this item: http://hdl.handle.net/10609/151216
Title: Aplicació de models de llenguatge extensos (LLMs) per a la generació de feedback de lliuraments avaluables amb rúbrica d’avaluació a l’educació superior
Author: Pérez-Milá Sánchez, Daniel
Tutor: Clarisó, Robert  
Abstract: Feedback is a very powerful influence on the learning process in higher education, making it essential to provide quality feedback to maximize its impact on academic performance. However, generating quality feedback requires a significant expenditure of time and resources. This often results in generalized, less personalized feedback, and in some cases, it is delivered with delay, thus diminishing its potential positive impact. For this reason, numerous studies have evaluated the use of artificial intelligence (AI) models based on natural language processing (NLP), revealing their potential to enhance the academic experience of students and relieve the workload of teachers. The emergence of large language models (LLMs) has opened the door to improving this aspect, although no definitive conclusions have been reached regarding their practical application using assessment rubrics. To address this challenge, in this work, we have integrated UOC's Canvas LMS with OpenAI's LLMs to generate specific and general feedback on student submissions in the Interface Design course of the Multimedia Degree, based on the activity assessment rubric. For this purpose, we have designed a prompt based on the analysis of characteristics that define good feedback. The results obtained show that by applying prompt engineering techniques, generative AI can produce feedback of notable quality, comparable to that of human teachers. Nevertheless, it is necessary to recognize that the limitations of the project may have influenced some discrepancies in the results obtained. Consequently, this study explores the potential application of this technology in future research to provide detailed feedback based on an assessment rubric in real educational settings.
Keywords: generative artificial intelligence
large language models
feedback
assessment rubric
Document type: info:eu-repo/semantics/bachelorThesis
Issue Date: 16-Jun-2024
Publication license: http://creativecommons.org/licenses/by-nc-nd/3.0/es/  
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
dperez_milaTFG0624memoria.pdfMemòria del TFG882,21 kBAdobe PDFThumbnail
View/Open
Share:
Export:
View statistics

This item is licensed under aCreative Commons License Creative Commons